-------
to
.0)
'G
c
.0
4w
fo
~ 03
cj 0
o "^
Is
O
00
Q_
U)
T3
CO
0
1
0
0
03
—1
en
ike Arrowhead Sewer Dii
to
— '
^
o
T3
CO
0
.c
5
0
k.
<
0
_*:
CO
Z
in
co
CO
Ij
00
5 co
iZ ||
O)
.§
0
E
to
Z
+->
c
CO
Q_
Owning Authority
>
.— . co
U «
O ^
Is
O
^|
OJ
n.
h-
•D
O
H—
E
03
CO
City of Stamford
i-
u
o
H-
£
CO
CO
O
i
1
Q.
h-
0
ir
*-•
c:
~
4->
Q.
0
Q
0
CO
>•
c
o
U
c
0
>-
_co
CJ
ei
-
4-"
D
O
D
c.
o
co
0
CO
"7
IT)
Q.
0
O
O
_0
0
O
o
0
O
M—
o
0-
Q.
~z.
0
o
O
c
0
0
^
_a • o
rly and used on a year-round basi
conomical alternative is preferred
low loading) or seasonally.
1 not handling routine loading.
• upgraded and has operated previ
.2 u ^ 2 o c >» C3 . O ff
5f o S « "8 - -3
" E 8 1 ^ 2 .2
g ca 'g S o tS •£
lfl|§l«
6, S — « BO o 3
° S -8 " I "§ ^3
•S e % c J5 3 -3
t/3 t« Cfl M ^ rt
E E 'g 'e E S |
S ^a ^ ^ S •«
n ii ^ i n
O Z « oo a! U ,
1-25
-------
Although MHFs were used for incineration of sludge long before the advent of the fluidized-bed
incinerators, they are falling from favor. Because of the increasing severity of the requirements im-
posed for pollution prevention, the off-gases product can no longer be released to the atmosphere
and substances that cause noxious odors must be decomposed at high temperatures before their
discharge. As a result, with the MHF, the off-gases must either be recycled to the hot region of the
incinerator or they must be led to separate burners operating at temperatures higher than 800 °C
(1,500°F). It is precisely this requirement that has been primarily responsible for the demise of the
multiple-hearth incinerator.
Few multiple-hearth incinerators have been constructed recently, at least in West Germany,
because the fuel costs involved in removing the odors from the flue gases often exceed the cost of
incineration. Economic factors favor using the fluidized-bed incinerator, which has a high enough
temperature to degrade the odors before emission to the atmosphere. The capacity of a multiple-
hearth incinerator, expressed in terms of the dry solids in the sludge, is only about one-third of that
of a fluidized-bed incinerator of comparable size.
Fluidized-bed incinerators in Germany are similar to U.S. designs. Controlled amounts of
dewatered sludge are fed into the fluidized bed of sand, which is heated to 750°-850°C (1,400° -
Table 1-9. Total production of nonstabilized wastewater sludge.
(In millions of cubic yards)
Country
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Netherlands
Norway
Portugal
Sweden
Switzerland
Spain
Turkey
United Kingdom
United States
1.8
3.4
3.1
58.9
98.1
—
0.5
31.4
5.2
1.8
—
6.3
3.8
—
—
48.4
160.9
Am.ount
Generated
(Year)
(1979)
(1977)
(1977)
(1980)
(1980)
—
(1980)
(1979)
(1979)
(1980)
—
(1978)
(1979)
—
—
(1980)
(1980)
Percentage
Incinerated
1.7
9.0
—
15.0
8.0
—
—
5.0
0.6
—
—
—
9.5
—
—
4.0
25.0
Cubic yards x 0.7646 = cubic meters.
1-26
-------
1,600°F). The constituent water evaporates, the combustible substances burn in the bed or
freeboard, and the combustion residues are swept out of the incinerator by the flue gases. The dust
content of the flue gases may be as high as 200 g/m3 (87 gr/cu.ft) and is reduced below the legally
permissible level of 100 mg/m3 (0.04 gr/cu.ft.) almost exclusively by electrostatic precipitators. In
some cases, wet scrubbers have been installed to eliminate the gaseous pollutants (e.g., sulfur diox-
ide, hydrogen fluoride, and hydrogen chloride). The operating costs for fluidized-bed incinerators
was about 300-500 DM ($114-$ 190) per ton of dry solids in 1983, including dewatering.
A combination of the multiple-hearth and fluidized-bed designs has been developed for the pur-
pose of direct heat-transfer drying of sludge and removal of noxious odors from flue gases. This
type of incinerator is equipped with a predrying and distribution zone that may have up to six
hearths; the fluidized-bed chamber is located underneath. Forty to sixty percent of the water contain-
ed in the waste material is evaporated in the predrying zone; the vapors flow through the upper sec-
tion of the combustion zone where all volatile components should burn out. Due to the highly effec-
tive drying operation, the needed cross section of the fluidized-bed zone becomes correspondingly
smaller.
To ensure the maximum burnout of the flue gases, an after-burning chamber may be linked to
this type of incinerator. In this type of plant, the combustion chamber may be kept even smaller
with corresponding savings in capital costs.
Rotary kilns are comparatively expensive to install, and the capacity of units of similar size is
much less than that of the MHF or FBF.
WASTEWATER SLUDGE INCINERATION IN JAPAN
Sludge incinerators were first installed in Japan in 1962. Table I-10 shows the number and type
for 1977, 1981, and 1983 and illustrates that the multiple-hearth type is predominant both in number
and in total incineration capacity. In particular, all incinerators with a capacity of 150 dry tons/ day
(t/d) or more are MHFs. The largest size is rated at 300 wet tons per day, based on loading rates
that are more conservative than those typical in the United States.
The percentage of total incinerating capacity accounted for by MHFs has dropped. In 1977 it
was 89 percent, but in 1981 it was down to 84 percent. The share represented by FBFs has in-
creased, and new types have recently been installed. These include wet oxidation, pyrolysis process,
and melting furnaces.
The first installation of a fluidized-bed incinerator was in the early 1970s. Most of these are in
smaller treatment plants that require intermittent operation due to low sludge production, and
capacities are up to 100 tons per day, wet cake basis. Most of the plants that plan to construct new
incinerators are smaller-size facilities, so the ratio of fluidized-bed units will grow larger in the
future.
Rotary kilns and moving-bed type incinerators are seldom considered the most suitable design
and are built only infrequently.
1-27
-------
Table 1-10. The number and types of sludge incinerators in Japan.
Multiple-hearth
Fluidized-bed
Inclined grate,
moving bed, and
others
Process devel-
opment units
(PDU)
Pyrolysis
Melting
furnace
Wet oxida-
tion
1977
No. of No. of Capacity
Units Plants (tons/d)
74 61 5,831
9 8 305
6 4 159
1981
No. of No. of Capacity
Units Plants (tons/d)
81 58 7,859
21 6 164
15 13 469
1983
No. of
Units
80
13
10
5
2
3
1-28
-------
CHAPTER II. IMPROVING SLUDGE INCINERATION METHODS
INCINERATOR DESIGN PRACTICES
This chapter utilizes case studies to discuss how several operational problems were overcome at
existing incineration facilities. The key to engineering these changes is the understanding of basic
design considerations, since these remedial actions began with what amounted to redesigns. It is
therefore necessary to present information required in the design of a typical incineration facility.
Although the discussion that follows presents the design procedures for a multiple-hearth incineration
process, many of the same concepts apply to other sludge combination processes as well.
Specifying the Feed
The first step in the design of a municipal sludge incineration system is defining the feed that
the system must handle. The usual parameters are:
1. Feed rate - stated in kilograms (pounds) of wet-basis cake per hour
2. Properties of the feed -
a. Percent solids (preferable to expressing as percent moisture)
b. Percent combustibles in the solids (volatiles plus fixed carbon)
c. Gross, or higher, heating value of the combustibles (HHV)
d. Ultimate analysis of combustibles
e. Presence of chemicals (e.g., lime) that react endothermically
Not often stated, but highly desirable, are softening and fusion points of the ash as determined
by ASTM Method D-1857-68 if a representative sample of the sludge or ash can be obtained.
In many instances, precise information on the feed is not known when specifications are
prepared. Instead, ranges of expected values are given to ensure that the furnace meets the needs of
the wastewater treatment plant (WWTP). All too often the designer increases this uncertainty when
communicating with the furnace manufacturer by specifying very wide and unrealistic ranges of
values for the feed parameters. The numerical permutations and combinations that result from this
practice prompt equipment designers to supply a single piece of equipment and expect it to operate
over an unrealistic range of conditions. It is perhaps analogous to specifying a car for use in carry-
ing a large family, pulling a camper on vacation and, at the same time, city driving, and getting 15
km/1 (35 mpg) fuel economy. The responsibility for making a "best guesstimate" on the sludge feed
and keeping this estimate within values that can be satisfied by a single-size piece of equipment is
clearly the designer's.
An effective alternative to specifying ranges is to specify various possible modes of plant opera-
tion and then develop, for each of these modes, the two major parameters mentioned previously,
feed rate and properties of the feed. One must then decide on:
• Minimum and maximum furnace exhaust temperature and
• Minimum percent oxygen in the exhaust gas or, in other words, the amount of excess air.
IM
-------
Prior to finalization of the specifications, heat and material balances should be prepared for each
case. A summary table should indicate, as a minimum, the following items:
1. Sludge combustion air requirement - mass flow rate and volume rate, usually in kg/hr (Ib/hr)
and 1/s (cfm)
a. Shaft cooling air recycle
b. Ambient air
2. Auxiliary fuel requirement - kJ/hr (Btu/hr) or in fuel volume terms
3. Auxiliary fuel combustion air requirement - units same as (1) above
4. Furnace exhaust flue gas volume - actual 1/s (cfm)
After preparation of the summary table that indicates minimum and maximum values for each
parameter, this table should be examined to determine if the adjustments required of the individual
equipment items are within the useful operating range of the equipment.
A single factor applied to the quantity of sludge to be processed should be the sole basis for
establishing the sizing of individual components. This will result in a harmonious design of all com-
ponents of the system.
Understanding the Combustion Process
Problems in the incineration of sludge solids cannot be solved without a thorough and complete
understanding of the combustion process. Although "hit or miss" approaches will occasionally yield
the desired results, they cannot be relied on to keep a MHF—or any furnace—operating properly.
The combustion control logic of a MHF is not understood by most engineers, and many furnaces in
place today have been improperly designed; in addition, information contained in some operating
manuals provided by manufacturers is inaccurate.
One approach to design is to consider that the MHF operates as a number of individual furnaces
connected in series. The mechanical design (i.e., size and number of hearths; size, number, and
location of burners; and size, number, and location of combustion air nozzles) and the combustion
control logic should reflect this consideration.
The heat and material system balances that have traditionally been used as a basis of design for
the MHF treat the MHF as a "black box." This is not to say that the First Law of Thermodynamics
is invalid. The answers obtained by this "black box" approach certainly represent overall fuel and
air requirements but do not give any clue to understanding the combustion processes occurring on
the individual hearths. Without this understanding, it is impossible to determine, for example, the
proper location of the auxiliary fuel burners. This usual approach gives the total heat required in the
furnace, but installing one single large burner somewhere in the furnace would usually not represent
an intelligent design. Additionally, the "black box" approach gives no clue as to the control loops
necessary for control of the furnace.
"THERMAL JUMP" REVISITED
In the early years of municipal sludge incineration, the theory of the "thermal jump" derived
from the work of Rudolphs and Baumgartner was used to justify a moderate exhaust temperature of
nominal 430 °C (800 °F) in the gases leaving the furnace. Their paper stated that "distillation of
II-2
-------
volatile matters from sludge containing 25 percent solids did not occur until 80 to 90 percent of the
moisture had been driven off, regardless of the temperature." Stated another way, starting with a 25
percent solids cake, volatilization should not occur until the total solids content of the sludge is be-
tween 63 and 77 percent.
More recent data indicate that self-sustaining combustion takes place when the mass reaches a
total solids concentration as low as 48 percent. For combustion to occur, volatiles must be driven
off, and, therefore, these data appear contradictory. This contradiction can be explained by a more
detailed examination of the phenomena taking place within the MHF.
As the sludge proceeds on its downward, serpentine path through the MHF, moisture is con-
tinuously evaporated. The sludge on the hearth develops furrows caused by the action of the rabble
teeth. When sufficient moisture has been evaporated, the very volatile parts of the sludge on the
upper ridge of the furrow begin to undergo destructive distillation before they are turned over by the
next pass of the rabble arm. The exact point at which this occurs is difficult to ascertain and is ob-
viously affected by many variables. The previously mentioned value of 48 percent total solids is not
an unreasonable estimate, however, if it is recognized that this is an average value for all sludge
within the hearth area, and the sludge on the upper ridges is substantially drier.
Ideally, to ensure complete combustion, the hearth where these volatiles begin to distill off
should have active combustion with visible flame and the hearth temperature should be 760 °C
(1,400°F) minimum. In addition, the volatiles should be exposed to this temperature for a defined
length of time. Unfortunately, this is not always possible to achieve in actual practice. Certain
organic materials called condensables may escape. The condensables and odors are the result of in-
complete combustion of volatile organic compounds. These compounds are products of pyrolysis and
materials distilled off from sludge during the drying process before the sludge reaches the active
combustion hearth. The condensables are the material normally caught in the liquid impingement
train of the EPA Method V paniculate test, the fraction often referred to as the "back-half catch."
A number of MHFs operating with exhaust temperatures in the 400 °C (800 °F) range have had
odor problems and have failed particulate emission tests because of the high contribution of the
back-half catch. These problems have been corrected by:
1. Substantially increasing the temperature of the combustion (hottest) hearth, which in turn in-
creases the temperature of the hearth immediately above the combustion hearth. This is
where volatilization is most likely to occur, and sufficient temperature is provided in the
gases to combust the distilled organics.
2. Providing an afterburner (either at the zero or top hearth or as an external unit), which
operates at sufficient temperature (nominal 760 °C [1,400°F]) to ensure complete combustion.
Of the two methods described above, the afterburner approach gives the greater degree of con-
fidence, especially where there are low boiling organics present. Both methods imply the use of aux-
iliary fuel to reach a higher temperature in the gases. When a dewatered sludge with an adequately
high total solids content is available, a 760 °C (1,400°F) exhaust temperature can be achieved
without the use of auxiliary fuel.
Furnace Operation
Incineration of sludge solids in an MHF occurs in four distinct zones (see Figure 1-3):
1. Moisture evaporation;
II-3
-------
2. Distillation and combustion of volatiles;
3. Combustion of fixed carbon; and
4. Ash cooling.
The boundary between zones may occur part-way across a given hearth. Assuming that the feed
rate and thermodynamic properties of the sludge are a given (i.e., not subject to control), there are
three primary variables that can be manipulated in a conventional-style MHF with all sludge combus-
tion air going to the bottom hearth(s):
1. Flow rate of sludge combustion air;
2. Auxiliary fuel firing rate; and
3. Rotational speed of the rabble arms (rpm).
Thus, the following parameters are the controlled variables of the combustion process:
1. MHF exhaust temperature (temperature of the uppermost hearth prior to the effect of a zero-
hearth afterburner or external afterburner);
2. Excess air in the MHF exhaust gas; and
3. Temperature of the combustion hearth.
Temperatures are usually measured by thermocouples, connected to appropriate instrumentation.
There is no instrument called an excess air meter. What is actually measured is volume percent ox-
ygen in the exhaust gas, either hot or cooled in a sampler. Measurement of hot flue gas containing
much water vapor from sludge moisture and combustion products is termed "wet" basis. If, on the
other hand, measurement is made on gases that have been cooled and scrubbed and have only
residual moisture at 100 percent relative humidity, it is termed "dry" basis. Future discussions refer
only to percent oxygen and, unless otherwise stated, this is on a "dry" basis. A handy reference
formula for converting percent oxygen (dry basis) to excess air is:
Percent Excess air = [O2/(21 - O2)] x 100
O2 = Percent oxygen (dry)
The temperature of the combustion hearth is henceforth referred to as the temperature of the "hottest
hearth."
A Look Inside a Multiple-Hearth Furnace
To fully appreciate why the conventional "black box" heat and material balances are inadequate
for the design and evaluation of the MHF combustion process, it is necessary to examine the com-
bustion process as it actually takes place. The results of a series of heat and material balances are
shown graphically in Figures II-1 and II-2. Since these graphs are for comparative purposes only, it
is advantageous to make certain simplifying assumptions.
• Assume shell heat loss is zero. Normally it amounts to only 2-3 percent.
II-4
-------
• Assume all shaft and rabble arm cooling air is recycled back to the MHF as sludge combus-
tion air. As a result there will be no net heat loss from the heated shaft cooling air being
discharged to the atmosphere.
• Assume zero percent combustibles in the ash. Normally this value would be 1-4 percent.
For these calculations, two differently conditioned and dewatered sludges were chosen. One
sludge was chemically conditioned (CCS) and the other was thermally conditioned (TCS). The ther-
modynamic properties of each follow.
CCS TCS
Percent total solids 25 40
Percent combustibles 65 60
Combustible heating value 23,000 26,000
(U/kg combustibles)
Combustible ultimate analysis
Carbon 50.64 53.73
Hydrogen 7.22 7.71
Oxygen 37.14 33.06
Nitrogen 5.00 5.50
Sulfur 0.00 0.00
TOTAL 100.00 percent 100.00 percent
In the graphs, the theoretical temperature of the products of combustion is plotted against per-
cent total solids at 9.0 percent oxygen (75 percent excess air). The conventional "black box" heat
and material balance would indicate, for the 25 percent total solids (TS), an exhaust temperature of
approximately 450 °C (850 °F). However, if it is assumed that combustion begins when the TS
reaches 50 percent, then the temperature on the hottest hearth (combustion hearth), where the 50
percent TS sludge is burning at 9.0 percent oxygen, should be approximately 980 °C (1,800°F).
Thus, what many have observed is shown graphically: the temperature of the hottest hearth is
significantly higher than the exhaust temperature. A temperature of 980°C (1,800°F) would probably
cause clinkers, and therefore the furnace operation would have to be modified.
Figures II-1 and II-2 also show the effect of operating at increased percent oxygen (excess air)
in the flue gas. It is commonly stated in the literature that MHFs are operated at excess air rates in
excess of 100 percent to assure oxygen for combustion. This large quantity of excess air is not re-
quired for complete combustion but is required to maintain the temperature of the hottest hearth at a
level that will avoid clinker ing or thermal stress to the furnace. Even though the desired result has
been achieved in existing MHFs, the simplified approach used heretofore has prevented an adequate
combustion control logic from being developed.
A value of 6.0 percent oxygen (40 percent excess air), under the proper combustion conditions
of time, temperature, and turbulence, is sufficient for complete combustion. In Figure II-1 the
temperature is plotted for values of 6.0 percent oxygen (40 percent excess air) and 0 percent oxygen
(0 percent excess air). The problems of excessive temperatures on the combustion hearth are obvious
in these situations relative to those with greater excess air. It is for this reason that where MHFs
have been operated at nominally 6.0 percent oxygen (40 percent excess air) a starved-air combustion
mode has been used.
II-5
-------
3000
2800
2600
2400
2200
2000
u. 1800
O
| 1600
a
| 1400
a>
1200
1000
800
600
400
200
65% Combustibles
10,000 BTU/lb Combustibles
I
I
3000
2800
2600
2400
2200
2000
1800
1600
1400
1200
1000
800
600
400
200
0
25 30 35 '40 45 50
Percent Total Solids
55
60
Figure 11-1. Heat and material balances using 23,000 kJ/kg (10,000 Btu/lb) combustibles
(°C = (°F - 32) 5/9).
Figure II-2 shows data for TCS. At 10.7 percent oxygen (103.3 percent excess air), the furnace
exhaust temperature is 760 °C (1,400°F), adequate to assure complete combustion and deodorization
without the use of auxiliary fuel. Because of the ballasting effect of the high excess air, the
temperature of the hottest hearth, at 50 percent TS, is approximately 870 °C (1,600°F), or only
110°C (200 °F) hotter than the exhaust temperature. Therefore, in a MHF with adequate air handling
capacity, a TCS can be easier to incinerate.
Hearth-by-Hearth Balances
To perform a hearth-by-hearth heat and material balance, it is necessary to have an extensive
data base to develop the "rate and heat transfer equations" that determine the success of any
mathematical model of this type. Some information on these has been published, but most MHF
manufacturers and knowledgeable consultants consider this information proprietary. Parameters that
should be included in a hearth-by-hearth furnace simulation model are:
II-6
-------
Q.
E
0)
1800
1600
1400
1200
1000
800
600
400
200
40
40% Total Solids
60% Combustibles
11,000 BTU/lb Combustibles
1
I
50 60
Percent Total Solids
1800
1600
1400
1200
1000
800
600
400
200
Figure II-2, Heat and material balances using 26,000 kJ/kg (11,000 Btu/lb) combustibles
(°C - (°F - 32) 5/9).
II-7
-------
1. Moisture evaporation rate (kg/m2/hr or Ib/ft2/hr) as a function of hearth temperature and gas
flow rate;
2. Heat loss to rabble arms as a function of temperature and the number of rabble arms on a
hearth;
3. Shell heat loss as a function of surface area and hearth temperature;
4. Air leakage through flap gate feeders and hearth doors;
5. Percent total solids at onset of volatile distillation and combustion; and
6. Separate combustion rates (kJ/m2/hr or Btu/ft2/hr) and heat release rates (Btu/fT3/hr) for both
volatiles and fixed carbon.
With a comprehensive model (incorporating a hearth-by-hearth heat and material balance), a sen-
sitivity analysis can be made for the wide variety of sludges that the furnace is likely to encounter in
a typical wastewater treatment plant, thereby making it possible to optimize the design and cost-
effectiveness of the MHF.
Moisture Evaporation Rate
With the exception of the fixed carbon combustion and ash cooling hearths, moisture evaporation
rate dominates the processes on all other hearths within the MHF. Moisture evaporation consists of
three steps:
1. Sensible heating phase — Before any evaporation can take place, heat must be added to the
sludge until the vapor pressure of the free water in the sludge exceeds the vapor pressure of
the water vapor in the flue gases. For typical sludge incineration in a MHF, the sludge will
reach approximately 71 °C (160°F) before rapid evaporation begins.
2. Constant rate phase — Once the above phase is reached, evaporation usually takes place at
a nearly constant rate over a certain range of moisture.
3. Critical moisture point — When the sludge has reached the critical moisture point, the dry-
ing rate occurring in the constant rate phase begins to fall. At this point (nominally 48-50
percent TS) the percent total solids in the upper ridges is higher than the average, and
volatilization of the combustibles begins to occur.
An optimum rabble arm speed is where the width of the level portion in the valley of the fur-
rows is approximately 3 cm (1 in). When rabble speed is too fast, this width will increase. When it
is too slow, it will fill in with sludge. Both of these have the effect of reducing the projected area
exposed to the hot gases and radiation from the roof.
When an attempt is made to "move the fire," or change the location of the hottest hearth by
slowing down the speed of the rabble arms, it can only be done at the expense of increasing the in-
ventory of sludge on a hearth, which can lead to a "runaway" (uncontrolled burning) furnace should
this large inventory of sludge begin to burn. Additionally, volatilization is likely to occur on a hearth
that is not up to proper combustion temperature (760°C [1,400°F] minimum), and unburned fuel
gases, including tars and oils, will be discharged from the furnace (observed as smoke).
II-8
-------
Attempts to control furnace operation by varying the speed of the rabble arms have largely prov-
ed unsatisfactory. Since a definite cause-and-effect relationship between rabble arm speed
(manipulated variable) and any other controlled variable has never been satisfactorily established,
rabble arm speed has been eliminated from the list of parameters considered in the hearth-by-hearth
heat and material balance.
Combustion Control Logic
The importance of correct combustion control logic for a MHF has been stressed. A MHF con-
trol circuit developed from incorrect hypotheses cannot succeed. MHF operators should not be
blamed for improper operation of their furnaces when they are not given the proper instruction
and/or control system.
In seeking economical MHF operation, there are a number of variables that should be
controlled:
• Temperature of flue gas leaving the furnace,
• Excess air (measured as percent oxygen),
• Temperature in the combustion hearth gases, and
• Location of the combustion hearth.
In a typical MHF fed at a constant rate, there are a number of conditions that can be
manipulated to help keep the controlled variables at their set points:
• Mass flow rate of sludge combustion air (recycled shaft cooling air plus outside air),
• Rotational speed (rpm) of the central shaft,
• Firing rate (temperature set point) of auxiliary fuel burners, and
• Hearth location of burners currently firing.
In a typical MHF, with almost all of the sludge combustion air introduced in the bottom hearth,
the combustion hearth is always the hottest hearth (HH). All hearths located above and below it are
at lower temperatures.
It is desirable that the temperature of the HH always be maintained at set point temperature.
Temperature control is achieved by varying the flow rate of sludge combustion air (SCA). In a
stable burning mode, the flow rate of the SCA is decreased in order to increase HH temperature.
Conversely, the flow rate of the SCA is increased in order to decrease HH temperature.
The maximum set point temperature for the HH is determined by either (1) the temperature, as
measured at the wall, at which the sludge begins to form clinkers, typically 870° - 980 °C (1,600° -
1,800°F) or (2) temperature limitation of the furnace, typically 1,000°C (1,900°F), which is based
on the grade of firebrick used and the alloy in the rabble teeth. A further discussion of slagging and
clinker formation is presented in Chapter III.
To maintain the burner flame safety circuit in a "purged" safety condition, a burner in the up-
permost hearth that receives auxiliary fuel is always lit. Unless it is needed to provide additional
heat to the furnace, this burner will remain on low fire.
II-9
-------
For safety reasons, all burners should be turned ON by the MHF operator and not by the con-
trol system. When the combustion control logic circuit determines that a burner should be turned
ON, a light on the control panel will signal the operator. The control panel lights, however, will
also tell the operator WHICH burner should be started. The control circuit will automatically turn
burners OFF.
When the combustion control logic circuit determines that more auxiliary fuel is needed in the
furnace, it will signal the operator as to WHICH burner to light. Once this is done, the combustion
control logic circuit varies the fuel firing rate until the desired results are achieved. The control cir-
cuit increases the fuel firing rate by AUTOMATICALLY increasing the burner set point temperature
on the "selected" fired hearth. Conversely, when less auxiliary fuel is required, the control circuit
AUTOMATICALLY decreases the burner set point temperature on the fired hearth, which in turn
decreases the amount of auxiliary fuel being fired into the furnace. This type of control loop is com-
monly called cascade control.
The maximum set point temperature for any fired hearth should be at least 50 °C (100°F) less
than the set point temperature of the HH. This avoids control circuit problems by assuring that the
control logic will not confuse a fired hearth with the HH.
DEWATERING SLUDGE
Many processes are available for removing water from sludge (dewatering) and thus preparing it
for combustion. Each of these processes can be designed in various ways, make use of different
commercial equipment, and be operated in alternative modes. Their objective is to ultimately pro-
duce a high-solids cake and thereby minimize auxiliary fuel usage in the subsequent combustion
process.
Thickening, conditioning, and dewatering are the common processes for removing water from
sludges. Other processes such as drying and dehydration are less common. Some processes (like
composting and combustion) result in water removal, but that is not ordinarily their primary
purpose.
Dewatering of sludge to produce a feed to the incinerator is a critical step for the process of
combustion. Both centrifugal and filter-type dewatering equipment have been greatly improved in re-
cent years. It no longer holds true that incineration is an unreasonable consumer of fuel. Many in-
cinerators today operate in an "autogenous" mode, using no fuel for moisture evaporation at all.
The burning quality or heat content of a sludge as fed to an incinerator may principally be im-
proved in three ways:
1. Remove water from the sludge more effectively by using the best available type of dewater-
ing equipment with the most appropriate conditioning process.
2. Before feeding sludge to the incinerator, dry the sludge partially or completely, in addition to
dewatering it, by using the heat from combustion that would otherwise be wasted.
3. Add a combustible material to either the dewatered or undewatered sludge as an augmenta-
tion of its heat value in relation to its moisture.
Dewatering Equipment
The most common approach is to design new facilities with the latest and most cost-effective
type of dewatering machine. In centrifuges, this is currently the variable-speed backdrive type that
11-10
-------
has reached the market in the past decade. In filtering equipment, the continuous-belt filter press has
taken a major share of the market once held by the vacuum drum filter. Another filtering method is
the recessed plate filter press, which was adapted from the chemical industry and is commonly pro-
vided in an automated version.
A strong endorsement of the belt press comes from the city of Hartford, CT. Similarly,
dramatic improvements in dewatering with the belt press were reported by the wastewater treatment
plants in Rochester, NY, and Duluth, MN. In these cases, the substitution was for rotary vacuum
drum filters. Some plants that have attempted to use continuous-belt filter presses in lieu of cen-
trifuges, however, have reported dissatisfaction with performance; examples of this are: Philadelphia
Southwest, Denver Metro, and Central Contra Costa in California. The principal problem with the
belt press arises when the sludge feed varies in quality and this causes a change in the required
polymer dose. The operator must be alert to such changes and modify the dose to meet the new con-
dition. Failure to do this results in the sludge being squeezed out at the sides and causes a severe
housekeeping problem. A centrifuge will give more solids in the centrate and a wetter cake in the
same situation but not cause a housekeeping problem.
Carver-Greenfield Dewatering (Drying) Process
The Carver-Greenfield (C-G) process was developed specifically for application to "liquid-
solid" slurries and has been successfully applied to various industrial slurries as well as industrial
and municipal wastewater treatment plant sludges. In the C-G process, water is essentially extracted
from sludge using a multiple-effect evaporator or vapor recompression dryer that results in a con-
siderable economy of steam compared to single-effect heat dryers.
A schematic diagram of the process as designed for the Hyperion Energy Recovery System
(HERS) in Los Angeles, CA, is presented in Figure II-3. Dewatered sludge is first mixed with an
oil, such as light-high-boiling solvent, which serves as a carrying or fluidizing medium. The carry-
ing oil assures that fluidity is maintained in all phases of the evaporation cycle and that formation of
scale or corrosion of the heat exchangers is minimized. Sludge-oil slurry is then pumped to a
multiple-effect evaporator where water is vaporized. The remaining solids-oil mixture is then cen-
trifuged and hydro-extracted (i.e., steam stripped) to separate the carrying oil from the solids.Carry-
ing oil is recycled for reuse in the evaporative cycle while the solids are removed for other pur-
poses, including subsequent combustion or reuse in agriculture. Oil and grease content (i.e., freon or
hexane extractables) of the Hyperion sludge varies from about 8 to 15 percent of the dry sludge
solid weight. These nonpolar components dissolve in the carrying oil. A sidestream of the carrying
oil is continuously withdrawn from the C-G evaporator and distilled to separate light fluidizing oil
from higher-molecular-weight sludge oils. Fluidizing oil is returned to the C-G process, and the
sludge oil is stored for subsequent combustion in the fluidized-bed gasifier. Startup of the Los
Angeles HERS System is scheduled for late 1985, with full operation in early 1986.
Continuous Belt Filter Press
Hartford, CT, Case History. Hartford, CT, began pilot testing belt filter presses in the spring
of 1978. Test results showed that significantly drier sludge cake was produced at a higher production
rate with a belt filter press (BFP) than could be accomplished with the existing vacuum filters. The
plant staff then conducted side-by-side performance tests of the best performing BFPs to select the
first BFP for procurement and installation. The first BFP was installed in 1979 and its startup and
shakedown were carefully monitored. Despite numerous mechanical problems and excessive
downtime (25 percent), the BFP quickly performed so cost effectively that approval for acquiring a
second press was granted only 4 months after installation of the first one. The payback period for
the first press was only 6 weeks. In selecting the second BFP, performance tests were again con-
11-11
-------
03
4—
c
0>
03
co ^
O ".
c/>
o
03
H—
M—
03
0)
03 O
> O
•z; 03
to £E
O >•
a en
f-s
£ .2
CO 03
P 03
N
CO O
CO
SO c
a> §3
o .i?
o w
Jr 03
a.-a
o w
'4-1 CO
c w
E w
03 03
^1 O
O O
W Q.
CO
03
D
O)
II-12
-------
ducted to evaluate overall performance, mechanical design, and maintenance features of competitive
presses and to incorporate the most desirable requirements into the bid specifications. The second
BFP was supplied by a different manufacturer than the first one and was installed in December
1979, just 8 months after the first one.
Since Hartford was one of the first plants to try the BFP, operational difficulties were expected.
Initial BFP operation included problems with the bearings, spray water pump, filter screen cleaning,
filter screen tracking, and filter screen seam closures. With assistance from the manufacturer, the
first press was retrofitted and upgraded for more reliable operation. The second through fourth BFPs
were from a different supplier and had fewer mechanical problems.
As more experience was gained, improvements were made in several key operating conditions.
Filter screen seam closure wearing was reduced by using scraper blades of a higher molecular
weight plastic, and an increase of from 500 hours to an average of 1,500 hours of filter screen
operating life resulted. Proper polymer conditioning of the sludge was a problem on all the BFPs. A
two-component liquid polymer mix was developed in experiments by a polymer supplier to reduce
dosage requirements to the same level as was required for the vacuum filters. Changes in the sludge
conditioning tank to improve polymer/sludge mixing also helped reduce dosage requirements and in-
creased operational flexibility for adjusting to the sludge's variable characteristics. Maintaining a
constant BFP feed by mixing the raw primary and waste-activated sludges from three plants requires
close operator control. Sludge blend variations of only 5-10 percent can cause a press screen plug,
resulting in sludge squeezing out at the ends of the rollers. This results in a reduction in cake solids,
lost production, and a messy cleanup job. In spite of these operating problems associated with reduc-
ing a new operating technology to routine production line practice, the operational improvements and
cost savings achieved with the BFPs at Hartford were dramatic.
Energy savings realized from the BFP conversion were significant. From the time of the plant's
startup in 1972, the activated-sludge mixed-liquor suspended solids (MLSS) concentration had
averaged 4,000-5,000 mg/1, requiring approximately 32.8 m3/s (100 million cubic feet per day) of
dissolved air. With the BFPs, the increase in dewatered sludge production has enabled the MLSS
level to be lowered to a more desirable 2,000 mg/1 range. The resulting decrease in the dissolved
oxygen demand reduced the daily air usage to approximately 18 m3/s (55 million cubic feet per
day). This reduction, in turn, reduced the electrical energy requirements of a 2,238 kW (3,000 hp)
air compressor by 20 percent, which amounted to a $200,000 per year savings in electricity costs.
Also, each vacuum filter had a 53.3 kW (71.5 hp) requirement as compared to 16.4 kW (22 hp) for
each belt press. This reduction in electrical use resulted in an estimated savings of $25,000 per year.
In addition, the elimination of the vacuum pumps resulted in a maintenance savings of $6,000 per
year. In total, these savings amounted to $231,000 per year.
The average specific fuel consumption or gallons of oil per dry ton (dt) sludge solids and the
moisture-to-volatile (M/V) ratio for the Hartford incinerator operations for the years 1978-81 are
shown in Table II-1.
The savings resulting from the belt filter presses are reflected in the sharp reduction in the
sludge cake M/V ratio, particularly in 1980 when the major fuel reduction was achieved. The net
reduction of an average of almost 0.34 I/kg (82 gal/dt) of oil would translate into savings of over
3.21 x 103m3 (848,000 gal) of oil at the 1982 dry ton production level of 9.41 x 106 kilograms
(10,351 tons). Coupled with the dramatic reduction in fuel consumption, there was also a 57 percent
gain in the volatile solids incineration rate per operating equipment hour, which is the key produc-
tion performance parameter. Furthermore, the average incinerator hours of operation per day for two
incinerators also dropped from 46.5 in 1978 to 35.7 in 1981, a 23 percent decrease. This meant that
only two of the three plant incinerators had to be used routinely.
11-13
-------
Table 11-1. Moisture-to-volatile ratio for Hartford incinerator operations.
Year Percent change
Variable 1978 1979 1980 1981 (1978-81)
Percent solids
Sludge cake M/V
Fuel consumption
13.8
8.6
125.2
14.5
8.1
116.1
18.5
5.8
60.5
19.5
5.4
43.5
+ 4
- 37
- 65
(gal/dt)1
Incineration rate 0.7 0.7 1.0 1.1 +57
(volatile tons/
incinerator/hr)2
\2
'gal/dt = 0.004 I/kg
2t/hr = 907.2 kg/min
These substantial results were accomplished after a considerable amount of time and effort was
invested by the Hartford plant management, staff, and operating personnel. The experience of Hart-
ford with the belt filter presses serves as a classic example of the opportunities that exist in many
plants throughout the country to achieve cost-effective performance by the adoption and modification
of new operating technologies.
Improvement of Centrifugal Dewatering by Steam Injection
Kansas City, KS, Case History. Municipal Wastewater Treatment Plant No. 1 is located
southwest of downtown Kansas City and situated on the Kansas River. The design flow is 0.3 m3/s
(7 mgd) with potential to expand to a maximum of 0.92 m3/s (21 mgd). Plant No. 20 is a complete
mix-activated, sludge-type secondary treatment plant with aerobic sludge digestion and sludge in-
cineration. The flow diagram for the wastewater treatment process is shown in Figure II-4.
Typically, secondary sludge alone will mechanically dewater to cake solids of 10-14 percent.
Following extensive testing and development work at Northwest Bergen County WWTP in
Waldwick, NJ, a centrifugal dewatering system was provided at Kansas City, employing steam
heating of the secondary sludge and an eddy current back-drive for the centrifuge. Steam was used
to heat the secondary sludge to 73.9°C (165°F) just ahead of the centrifuge. The eddy current back-
drive for the centrifuge automatically adjusts the differential speed between the bowl and the scroll
to maximize solids' residence time in the centrifuge and to keep the centrifuge operating at full load
regardless of percent sludge solids in the incoming feed. It is not known if this sludge heating with
steam can be universally applied to all secondary or activated sludges. It certainly worked well at
Kansas City Plant No. 20, as shown in Table II-2. The fire tube boiler is designed for an operating
pressure of 700 kPa (100 psig). It is normally operated at 400 kPa (60 psig), as this is all that is re-
quired for steam injection into the sludge feed line to the centrifuge. Also, operation much below
400 kPa (60 psig) (saturated steam 153°C [307 °F]) is not recommended because of potential corro-
sion problems at lower steam pressures and temperatures. The steam requirement at Kansas City
Plant No. 20 for the conditions specified in Table II-2 and Figure II-4 is 1,114 kg/hr (2,455 Ib/hr)
of 400 kPa (60 psig) saturated steam. The fire tube boiler is rated at 1,800 kg/hr (4,000 Ib/hr), so
some hot gas is bypassed around the boiler to balance the steam supply with steam demand.
11-14
-------
Steam
Steam
Thickened
Secondary Sludge
Waste
Heat
Boiler
Shutoff
Valves
MB
Water
Venturi *T^
Scrubber k
Thickened
Primary Sludge
Fluidizing
Air Blower
Dorr-Oliver
Stamford, CT
To Turbo
Generator
Kaw Point
Plant No. 1
Only
Stack
Water
Tray
Cooler
Effluent
Figure II-4. Municipal Wastewater Treatment Plant No. 1, Kansas City, KS.
DRYING SLUDGE BEFORE COMBUSTION
Drying the sludge before incineration has not been adopted in the United States, except where
cocombustion of sludge and solid wastes is the objective. Stamford, CT, has done this successfully
in a rotary dryer for several years (see Chapter III). Attempts at Harrisburg, PA, to use a "porcu-
pine" dryer have been unsuccessful, although this unit has worked satisfactorily in Europe. At the
Flint River WWTP in Clayton County, GA, the Heil dryer is used and the sludge is pelletized.
Two examples of systems for combustion sludge that is burned by itself are located at plants in
Minneapolis-St. Paul, MN, and Norwalk, CT. At the Metro WWTP serving the Twin Cities, two
rotary dryers were provided in the facilities started up in 1983 to provide a dry sludge option in case
a potential offsite use was developed or if very wet feed was being delivered from the sludge
dewatering system. As of this time, these dryers have not been used. Because thermally-conditioned
sludge is developed very effectively, the main operating problem has been too "hot" a feed to the
furnaces, instead of excess moisture. At the Norwalk plant, an add-on dryer system of the fluidized-
bed type was started up in 1983 and is reported to be operating successfully. This system links the
previously installed Fluosolids® combustor with the new dryer by lifting hot sand to the dryer,
where the feed sludge becomes mixed with the sand. The now-dried sludge solids mixed with sand
are dropped back to the main combustion chamber and burned. A further linkage of the two vessels
is that the fluidizing air for the dryer is heated in the second stage of the air preheater that extracts
energy from the hot combustion gases. The first stage provides a hot air stream to the windbox of
the combustor. The benefit of such integration is savings in fuel. The water is taken away from the
11-15
-------
Table 11-2. Improving dewatering by stream injection-
Municipal Wastewater Treatment Plant No. 20, Kansas City, KS.
Secondary Sludge Heating
Sludge feed
Secondary sludge - Percent solids in centrifuge cake
Primary sludge - Percent solids in centrifuge cake
Ratio of primary solids/secondary solids
Yes
23
27
1.7 to 1
No
12
27
1.7 to 1
Percent solids in composite centrifuge cake
(feed to combustor)
Fluosolids® combustor
Capacity - Pounds feed solids per hour
Auxiliary fuel - Btu x 106 per ton solids
Power - kWh per ton solids
Operating costs (47.4 tons/solids per week)
Operator hours/week - shifts/week
Operator labor cost ($25,000/year/person)
dollars/ton solids
Auxiliary fuel cost ($5/million Btu) -
dollars/ton solids
Power cost (50 per kWh) - dollars/ton solids
Total labor, fuel, and power
Savings per ton of solids
Annual savings (2,275 tons/year solids)
25
1415
4.18
290
72-9
19.78
$20.90
$14.50
$55.18
$33.00
$75,000
18.5
1079
8.56
380
96-12
$26.37
$42.81
$19.00
$88.18
1 kg = 2.20462 Ib 1 kJ = 0.948 Btu 1 Mg = 1.1023 ton
sludge solids at a much lower "cost" in calories per gram (Btu's per pound) of water than it would
be in the combustor. The moist off-gas leaves the dryer at a much lower temperature, perhaps
100° - 130 °C (220° - 260 °F), than it would if the drying was being done in the main combustor.
This moist, odorous gas is then wet-scrubbed to remove its moisture burden before being routed
back through the first stage of the preheater to the main combustor windbox and into the combustor,
where any remaining odors are destroyed.
11-16
-------
Another major benefit of the Norwalk installation, as there would be with any predrying pro-
cess, is that the capacity of the combustion device is expanded greatly. Much larger amounts of
sludge solids can be burned per hour because the limitation caused by necessary drying in part of the
combustion unit is no longer present. At Norwalk the burning capacity was doubled. Thus, a plant
with a sludge disposal load in excess of its burning capacity might find that adding a predrying step
expands its capability and reduces fuel consumption per unit of solids handled at a much lower cost
than if it elected to add another combustion system.
Drying ahead of combustion, of course, is not new. It has been practiced at Allegheny County
Sanitary District's plant (Alcosan) in Pittsburgh, PA, for many years, as well as in other locations
with Raymond-type flash dryer equipment that is linked to a combustor.
Energy-Efficient Dehydration Prior to Combustion
Heat drying to remove moisture prior to combustion, as discussed in the previous sections, can
be accomplished in a number of different ways. It can be thermodynamically advantageous provided
that the moisture is removed with less energy than that required to accomplish the same thing in a
furnace. Two heat drying processes are examined here: indirect contact steam dryers and the C-G
multiple-effect evaporation process. Other drying systems are available but generally require greater
energy input than the two examined here.
An energy balance for indirect steam drying shows that the thermal requirement for drying is
about 2,910 U/kg (1,250 Btu/lb) of water removed, which equates to 99,000 MJ/day (94 MBtu/day)
if it is assumed that the sludge is dried to 80 percent solids. Energy recovery from the furnace ex-
haust is about 77.2 percent of input, or about 104,430 MJ/day (99 MBtu/day). Recovered energy
essentially balances the requirement of the steam dryers. Net energy cannot be recovered unless the
cake solids concentration is significantly increased above the 20 percent assumed.
A four-effect C-G process will thermodynamically operate at about 930 kJ/kg (400 Btu/lb) of
water removed and produce a product with about 1 percent moisture. An energy balance for C-G
drying and combustion shows that 24,000 MJ/day (32 MBtu/day) is required for the C-G process
and about 111,000 MJ/day (105 MBtu/day) is recovered from the combustion system; thus, signifi-
cant net energy production can be accomplished even with 20 percent sludge solids.
Energy-efficient dehydration of wet sludge cake can produce a material capable of autogenous
combustion without the need for supplemental fuels. Sufficient steam can, in some cases, be
generated from thermal processing for net electrical power production as well as operation of the
drying process. Importation of fossil or alternative fuels to the treatment plant site is not required,
and the technology involved appears to have a low odor potential compared to other alternatives. For
these reasons, energy-efficient drying of digested, dewatered sludge using the C-G process was
selected by the city of Los Angeles.
Other Drying Methods
In addition to the use of hot gases from the incinerator in external direct or indirect drying
devices, other interesting developments in promoting more efficient drying within the incinerator
have taken place in Germany and Japan.
In Germany, several "hybrid" furnaces have been built that combine the fluidized-bed and the
multiple-hearth configurations in one shell. The MHF portion is above, and the combustion region is
below. Partially dried cake drops into the fluidized bed and is burned with minimal fuel demand.
Also, the diameter of the fluidized bed is lowered by this design because it is not evaporation-
11-17
-------
limited. Two such Lurgi furnaces have been in service in the Frankfurt treatment plant for several
years.
In Japan, studies have been carried out in experimental systems where gas is recirculated
through the drying zone to increase velocity over the sludge and increase the rate of evaporation.
This was found necessary in units operated in the starved-air combustion (SAC) mode because of the
low excess air supply, which in turn was intended to minimize the oxidation of chromium to the
hexavalent state, the most toxic form. One or two of the hearths are actually operated in reducing
conditions for this purpose, which in turn mandates the lesser air supply. This technology is ex-
pected to become increasingly popular in Japan and those parts of the United States with sludges
high in trivalent chrome.
FUEL NEEDS
Stable operation of a sludge solids combustor requires an equalization between heat in and heat
out; this is known as a heat balance. Heat in can be in the form of sensible heat energy and in-
troduced in the air supply, which is at an elevated temperature, or in the combustible material. Heat
out can be as sensible heat energy contained in combustion products, latent heat of water vaporiza-
tion, radiation and surface losses, and heat that is lost in excess air and equipment cooling. The
balance point is dependent on the amount of excess air that is allowed to pass through the furnace
and the recovery of energy from the gases leaving the furnace. A system requiring no heat other
than what is provided by the sludge cake is termed autogenous.
A furnace being fed sludge cake that does not have sufficient fuel value to balance the heat leav-
ing the system requires auxiliary heat to stay in balance. This is supplied by burners using commer-
cial fuel such as natural gas or fuel oil, which is fired into the furnace to inject heat.
Sludge cake that has an excess of heat content, as compared with the heat needed for drying and
elevating the combustion gas stream to target flue gas temperature, is termed superautogenous. If
superautogenous burning is conducted at high temperatures, equipment damage and ash fusion can
occur. Control of the temperature can be achieved by varying the air supply or by modifying the
cake moisture.
Fuel is not used in a sludge incinerator to burn sludge—it is burned to evaporate water. Any
sludge cake will burn by itself once it gets dry enough. Fossil fuel is only used to remove most of
the water, but sludge solids do not have to be bone dry to burn; self-sustaining combustion will
occur on a sludge lump once it gets to about 50 percent moisture or less as was discussed in an
earlier section. Frequently, dewatering sludge to a higher solids content is more economical than
using fuel in the incinerator to evaporate water.
Fuel is also needed for other reasons:
1. Startup heating and holding at standby of a furnace requires fuel that cannot be furnished by
sludge combustibles. Cooldown may require some fuel to manage the rate of temperature
drop.
2. Safety standards may require that at least one burner is firing at all times.
3. If a mandatory exit temperature is required by air quality permit or policy of plant manage-
ment, some fuel may be required to assure compliance with this requirement.
Heat sources that can be used to supplement the sludge combustibles to achieve heat balance at the
desired furnace exit temperature and oxygen levels are found in Table II-3.
11-18
-------
Table 11-3. Supplemental energy sources.
Source
Comment
Natural gas
Fuel oil
Air
Scum
Coal
Waste oil
Refuse-derived fuel (RDF)
Woodmill waste or wood chips
Paper mill waste
Industrial oily wastes
Repulped paper
Sometimes only available on an interruptible basis;
normally the lowest cost fuel
More expensive than gas
"Hot" air from the MHF shaft cooling system exhaust;
preheated air from the fluidized-bed windbox
Requires a special concentrator for waste removal and
a metering pump for controlled feeding
Requires dustproof construction within the plant for
handling; must use gas or oil for MHF temperature
control
Practical if a source of reasonable, steady volume is
available. Concern must be exercised regarding handl-
ing and contaminants
A potential fuel; has not yet been successfully used
Possible handling problems (bridging in bulk bias);
wood chips have been used successfully in FBFs
May contain a high ash level due to filler
Not recommended because of potential toxic emis-
sions and corrosion to ducts and scrubber system
Water on the fibers may nullify heat gain from
cellulose; requires dewatering and mixing with sludge
Understanding Incineration to Minimize Fuel Needs
Fluidized-Bed Incineration at Duff in Creek, Toronto. On the north shore of Lake Ontario,
east of metropolitan Toronto, the first stage of one of Ontario's largest sewage treatment plants is in
operation. The Duffin Creek Water Pollution Control Plant is being built in four stages. Stage 1 has
a daily treatment capacity of 289,250 m3 (50 mgd). The sludge combustion and energy recovery
system at this plant is a set of parallel trains of fluidized-bed combustors, each rated at 67 X 106
kJ/hr (63 x 10" Btu/hr).
This combustion facility has several interesting and innovative features. The fluidized-bed reac-
tors are designed to operate either as CWB or HWB units. The exhaust gas-to-air heat exchanger is
piped so that it can be bypassed or so that a part or all of the reactor hot gases flow through the
11-19
-------
heat exchanger. This means that the fluidizing combustion air to the reactor can be controlled from
about 60 °C (140°F) (heat developed by compression in the blower) with no air preheating when
burning drier sludges, up to a maximum air preheat of 621 °C (1,150°F) when burning wetter
sludges. Part of the steam generated in the water-tube waste heat boiler is used in a turbine to drive
the fluidizing air blower. The balance of the steam is used for process steam and for building
heating. In the future, when three more stages are added and plant capacity has increased, turbo
generators will be added to produce electrical energy that will be used in the wastewater treatment
plant. Venturi scrubber effluent with collected ash is thickened and then dewatered. Thickener
overflow is recycled to the Venturi scrubber. Dewatered ash is blended with dry ash from the waste
heat boiler dust hoppers to produce a damp ash material that is trucked to a landfill.
Successful performance testing was carried out on the system in January 1982. Air quality test
data and operating conditions during testing showed that total hydrocarbons in the stack ranged from
2 to 7 ppm and averaged 4.5 ppm; stack particulates were 0.13 gm/kg (0.25 Ib/ton) of feed solids;
combustibles in the ash were less than 1 percent. To verify the adequacy of the installation, higher
solids cake was imported from the nearby Lakeview plant. This thermally conditioned sludge filter
cake was 34-36 percent solids, and the performance was satisfactory; no fuel was required.
In early 1983, the Duffin Creek Plant was operating at reduced capacity. The belt filters were
not producing the projected 30 percent cake solids but rather cake solids ranging from 17 to 27 per-
cent and averaging about 20 to 22 percent. Also, the amount of volatile solids in the sludge was
below the design level. Because of the higher water content of the sludge and the lower volatile
solid content of the sludge solids, it became desirable and necessary to maximally preheat the air to
a temperature of 621 °C (1,150°F) so as to minimize the auxiliary fuel requirement. With increased
air pre-heating, the gas temperature to the waste heat boiler was lower, resulting in reduced steam
production. When burning a 22 percent solids concentration sludge of which 70 percent of the solids
are volatile, the auxiliary fuel requirement goes from 0 to 0.14 liters of No. 2 oil per kg of solids (0
to 34 gal/ton). The combustor capacity drops from a design of 4,350 to 3,570 kg/hr (9,600 to 7,860
Ib/hr). Steam production drops from a design value of 11,750 to 10,054 kg/hr (25,906 to 22,165
Ib/hr). The fluidizing air blower consumes the same amount of steam, 4,470 kg/hr (9,850 Ib/hr) in
both cases, but in terms of percentage of total steam generated, steam usage goes from 38 percent of
the total to 44 percent of the total. Tests were carried out with other belt presses for sludge dewater-
ing. These newer units in tests produced 4 to 5 percent higher solids content than is being obtained
with the existing units. These drier cakes, coupled with increased sludge generation rates, reduced
the usage of auxiliary fuel and thus reduced operating costs, but this was not considered sufficient.
In October 1983, purchasing commitments were made for membrane-type automatic plate-and-frame
filter presses..
UPGRADING EXISTING MHFs TO REDUCE FUEL NEEDS
Of the various types of combustors in common use, two offer the opportunity to make the dry-
ing process more efficient and two do not. Those that do are the MHF and the IEF. Both the MHF
and the IEF configurations provide a long residence time for the solids and permit separation of the
drying and burning zones for stagewise process control, a feature that is not possible in a FBF and
difficult in a rotary kiln furnace.
A major improvement in using the MHF as a better drying device came through a study at the
city of Indianapolis's Belmont Street Treatment Plant in 1981. This showed that substantial savings
in fuel consumption could be achieved with an appropriate process control strategy. The fundamen-
tals of this strategy are to (1) keep the burn zone low in the furnace, about two hearths above the
bottom; (2) utilize shaft cooling exit air to the maximum degree; and (3) minimize the excess air
11-20
-------
amount and off-gas temperature as much as possible while still achieving desirable air quality stan-
dards. The first and second goals provide the maximum amount of area above the burning sludge for
the drying process and use heat generated by the sludge combustibles instead of purchased fuel. The
third goal prevents heat waste by minimizing mass flow and heat content in the off-gas. These main
goals, together with supporting instructions, were also the basis of revisions implemented in sludge
incinerators at Nashville, TN; Buffalo, NY; and Hartford, CT, all of which are subautogenous feed
situations. The following list is a blend of general operating procedures utilized at Indianapolis and
Nashville:
1. Maintain steady feed rate.
2. Maintain top hearth draft at 0.51 to 2.03 mm (0.02-0.08 in) of water negative.
3. Maintain oxygen level at 4-5 percent at full rated loading, somewhat higher at partial load,
and a maximum of 8 percent at 55 percent of rating. These are for readings in-situ on hot
gases containing vaporized and combustion moisture; after scrubbing and cooling, the values
would be 2-3 percent higher because most of the moisture would be removed.
4. Control burn zone on hearth 6 (of 8, Indianapolis) or 7 (of 10, Nashville). This is managed
mainly by burner firing rate control.
5. Minimize or eliminate any use of burners on hearths above the burn zone. If a choice is
made, use lowest burners and only to control the initiation point of combustion.
6. Keep center shaft speed at a low, steady rate—100 seconds per revolution or longer. Do not
attempt to manage burning zone by varying the shaft speed.
7. Use only shaft cooling return air for sludge combustion.
8. Limit maximum hearth temperature to 843° - 899°C (1,550° - 1,650°F) to prevent formation
of clinkers.
9. Allow top hearth temperature to be what it will; do not use higher hearth burners to reach an
arbitrary standard, although occasional use for smoke abatement during upset conditions is
allowable.
As an added benefit, this strategy is reported to provide improved performance of scrubbing
equipment and better compliance with emission standards.This result is probably due to the lowered
air velocities attributable to less excess air and lower off-gas temperatures, which cause less entrain-
ment of fly ash.
Hartford, CT, Case History
The Hartford Water Pollution Control Plant provides primary and secondary wastewater treat-
ment for more than 170 X 103m3 (45 Mgal) of wastewater per day and generates in excess of 180
Mg (200 tons) of sludge cake per day. The sludge handling facility was originally designed in 1968
with four dissolved air flotation thickeners, five vacuum filters, and three multiple-hearth inci-
nerators. In 1978, before the conversion to belt filter presses described previously in this chapter,
the vacuum filters averaged 13.8 percent cake solids. Production required continuous operation of
three of the five vacuum filters, with two of the three incinerators operating around the clock. The
plant operation experienced the typical production and maintenance problems associated with handl-
11-21
-------
ing an extremely wet sludge cake. In addition, the Hartford plant started to receive sludge from
satellite plants in East Hartford and Rocky Hill. The incinerator operations were plagued with the
operating problems of handling very wet sludge cake and consuming large amounts of fuel. A Hart-
ford incinerator is shown schematically in Figure II-5. It is equipped for either gas or oil operation
and rated at 11.3 wet Mg (12.5 wet tons) per hour. No common operating procedure was used by
the incinerator operators. Each operator had certain specific practices and techniques for maintaining
Sludge Feed
Gas/Oil
Burners (1)
Burner Air
Cooling Air Exhaust
"Pop"
Damper
Return Damper
Cooling Air
8
11
Return
10
T
Ash
Exhaust
(1)4 Burners/Hearth
(2I1 Damper/Hearth
Rabble Arm
Cooling Air
Auxiliary Air
Figure II-5. Hartford incinerator system schematic.
11-22
-------
temperatures on various hearth levels and for managing incinerator airflow. The operators' preoc-
cupation with just burning the very wet sludge cake resulted in many inefficient operating practices.
Fuel Reduction Program. Reductions in fuel consumption for sludge cake incineration were
accomplished in two ways. The first was to substitute belt filter presses in place of coil-type vacuum
filters, and the second was to modify the MHF operating methods and upgrade their control system.
The second change involved the acceptance and implementation of advice presented by a specialized
consulting firm. Based on a careful study of equipment and operating practices, a new operating
mode was undertaken, and the collection and analysis of data was improved. The result was more
uniform procedure among the various shift operators and adherence to a greatly revised format of
operating variable control. Overall, this reduced fuel consumption by 83 percent, which in dollar
terms amounted to a savings of over $1,300,000 for 1982 compared to operation in 1978 and before
the changes.
To provide an accurate baseline for comparing the fuel reduction achieved by converting to belt
filter presses and improving incinerator operation, a statistical analysis was made of key performance
data, including temperature and oxygen content of the off-gas, fuel consumption, and air usage for
past operations and each of the years during which changes were made. In addition, the correlation
of specific fuel consumption (SFC) measured in liters of oil consumed per dry kilogram (gal/ton)
with the absolute sludge cake moisture to volatile ratio by weight (M/V) was computed to provide a
more comprehensive measure of change for comparison:
M/V
s x v
where
s — fraction solids in cake
v = fraction volatiles in cake solids
The M/V ratio is a site-specific parameter that is useful at a given plant for process control, but
is limited for interplant comparison to those having similar volatile heat values and running at com-
parable excess air ratios and breech temperatures. The average specific fuel consumption for the
Hartford incinerator operations in 1978 was 0.52 liters of oil per dry kilogram (125 gal/ton). The
sludge cake solids averaged 13.8 percent and the volatiles averaged 77.1 percent. The sludge cake
M/V ratio, which is directly related to and principally determines the specific fuel consumption de-
mand, averaged 8.6; this is relatively high. For example, if evaporative effectiveness is 4,700 kJ/kg
(2,000 Btu/lb) of water and sludge volatiles have 23,000 kJ/kg (10,000 Btu/lb) of heat value, both
expressed in gross heating value, cake having an M/V ratio of 5.0 would be autogenous. Any higher
M/V in this plant would require fuel. At 8.6, the cake would require 3.6 times 4,700 kJ/kg (2,000
Btu/lb) or additional 17,000 kJ/kg (7,200 Btu/lb) of volatile solids to reach balance at the exit gas
temperature and oxygen content because of the large amount of moisture in the sludge cake.
Operational Testing and Analysis. An operational analysis was made of the Hartford in-
cinerator operations and included measurements of airflow; analysis of exhaust gas; and assessments
of key instrumentation and controls, existing operator-specific practices, feed rate management,
airflow management, burner use profiles, hearth temperature profiles, and combustion zone location
and control. A demonstrated and proven kinetic incinerator analytical model was also used to deter-
mine the optimum loading rate and plant operating mode that would result in the minimum possible
fuel consumption. Preliminary investigation of operator practices found that no uniform operating
11-23
-------
procedure was used. There were several operating practices and lack of controls that were con-
tributing to excessive fuel consumption such as:
• Combustion occurring too high in the incinerator;
• High exhaust gas temperatures;
• High draft settings and too much auxiliary air;
• Misuse of heated rabble arm cooling air;
• Improper burner use choices;
• Improper techniques for controlling combustion zone location; and
• Lack of remote operator controls for airflow dampers and burners.
A preliminary analysis indicated that optimum airflow management could result in a potential
fuel reduction of 70 percent when burning sludge cake with an M/V ratio of 5.0 at an incinerator
loading of 6 tons/hr. Two examples of M/V at 5.0 would be 22 percent solids cake at 71 percent
volatile and 25 percent solids at 60 percent volatile. The two main parameters of furnace off-gas
(breech temperature and oxygen in the off-gas) must be kept as low as possible without violating air
emission limits.
The kinetic analysis for these conditions predicted that the potential fuel consumption for the
Hartford operation with such a dry cake was zero. This analytical result agreed with the empirically
based preliminary estimate drawn from airflow management, since an additional 30 percent of fuel
savings could be reasonably expected from improved combustion zone location control, optimum
burner use, improved load rate management, and the synergistic effect of these operating mode
techniques on fuel consumption. Based on these results and those from similar programs in In-
dianapolis, Buffalo, and Nashville, periods of autogenous combustion were expected with the new
operating mode. Autogenous combustion was achieved several times during the operational trial and
demonstration test for as long as 8 hours. During the period of routine use of the new operating
mode, there were many days in which no fuel was used over a 24-hour period. Based on the opera-
tional trial tests and analyses, a new operating mode with specific instructions and operating settings
was developed. The new operating mode was then demonstrated in full-plant operation for a 2-week
performance demonstration test period. On-the-job operator training in the use of the new mode was
also accomplished at the same time, since this is the only way to ensure continued good operation.
After completion of the successful performance test, the operating mode was further refined for
routine operational use.
The new operating mode was characterized by the following general operating guidelines:
• Maximize the use of the heated rabble arm cooling air return;
• Use the least possible draft (i.e., just slightly negative) to minimize air leakage;
• Combust on the third lowest hearth to maximize drying area;
• Replace cold auxiliary air supply with heated cooling air return;
• Minimize excess air by observing and minimizing oxygen in flue gas;
11-24
-------
• Use lower hearth burners to maximize drying temperature;
• Eliminate airflow to top hearth burners;
• Control combustion zone location with burner use profile;
• Slow center shaft speed to improve sludge drying;
• Discontinue use of hearth No. 5 burners; and
• Use the following operating parameters in breech:
Oxygen 4-5 percent in the raw off-gas, which equates to about 7.5 percent on a
moisture-free basis after scrubbing
Draft 0.05-0.20 cm (0.02-0.08-in) w.c. (negative)
Temperature 370° - 425 °C (700° - 800 °F)
The specific operating instructions that constituted the new operating mode were given to the in-
cinerator operators and included procedures for sludge load management, incinerator operation con-
trol, specific settings for normal operations, combustion zone location control, standby and startup
operations, and techniques to control sludge cake "burnouts." The most effective incinera-
tion/dewatering configuration was for two BFPs to feed each operating incinerator. The optimum in-
cinerator loading rate was found to be 6.6 Mg (6 wet tons) per hour per incinerator, approximately
half the design basis, based on analysis and trial tests of load rates between 5.0 and 7.7 wet Mg (4.5
and 7 wet tons). The 6.6-Mg (6-ton) per hour load rate was the minimum rate needed to keep up
with the overall plant production rate and still minimize fuel consumption, considering that the
average sludge cake M/V ratio was 4.5. (A 4.5 M/V would mean, for example, 22 percent solids
and 79 percent volatile or 25 percent solids and 67 percent volatile.) The improved operating mode
also enabled a further reduction in the M/V ratio because the new mode allowed the dewatering
presses to be slowed down, resulting in drier cake.
Fuel Reduction Results. The new incinerator operating mode was placed into routine operation
immediately following the 2-week performance test conducted in January 1982. Operational data for
1982 were analyzed to measure and compare the fuel reduction achieved. Shown in Figure II-6 is
the computed least squares correlation of the average specific fuel consumption vs the sludge cake
M/V ratio for the baseline period 1978-81 and for 1982. The improved thermal operating efficiency
achieved is reflected in the change of the slope of the relationship. This result was quite similar to
what occurred in Indianapolis, Nashville, and Buffalo when these plants implemented similar
operating techniques. Figure II-7 shows the average specific fuel consumption for the Hartford
operations from 1978 through 1982. The average specific fuel consumption for 1982 was 0.091/dry
kg (21.1 gal/dry ton) as compared to 0.18 (43.5) for 1981, approximately a 51.5 percent reduction.
With this improvement, the total fuel reduction achieved by Hartford between 1978 and 1982
amounted to 0.43I/dry kg (104 gal/dry ton), or 83 percent, which at the 1982 production level
represented a savings of 4,074.568 m3 (1,076,504 gal) of No. 2 fuel oil as compared to 1978.
In addition to reducing direct fuel consumption, the new operating mode provided increased
operating flexibility with the equipment because the incinerators could now be operated efficiently at
load rates 50-60 percent of capacity, which was not possible before without paying a tremendous
penalty in excess fuel consumption. Incinerator operation is also now characterized by cooler max-
11-25
-------
125
100
c
•- "c
0.^0
3 >•
O —
06
0) tr
3 O
H- «
£|
'o ra
75
25
I
I
45678
Sludge Cake Moisture to Volatile Ratio (Ib/lb)
Figure II-6. Specific fuel consumption vs sludge cake M/V ratio before and
after incinerator operating mode change at the Hartford plant.
imum operating temperatures, more steady state control, less particulate emissions, and reduced
maintenance on internal incinerator parts.
Cost Savings. The nominal cost savings from reducing incinerator fuel consumption on an an-
nual basis were estimated from the change in the specific fuel consumption from 0.52 to 0.091/dry
kg (125 to 21 gal) of oil/dry ton. Based on 1982 production of 9,388.4 dry Mg (10,351 dry tons),
the savings would be over $1,076,000 per year using an estimated price of $0.26 per liter ($1 per
gallon) for No. 2 fuel oil. The total estimated annual operating cost savings from converting to belt
filter presses and the new incinerator operating mode are over $1,300,000 per year.
AUTOGENOUS COMBUSTION
In Japan
In Japan, there are at least two plants utilizing the principle of autogenous combustion with
MHFs. For dewatering, a belt filter press is used at the Yotsuya Treatment Plant in Takaoka City,
while a recessed plate pressure filter is used at the Hojin Treatment Plant in Nagoya City.
Autogenous combustion at these plants was made possible not only by such factors as optimum
dewatering and incineration processes, but also by the improvements that resulted from the challenge
to operators and other personnel to achieve autogenous combustion.
11-26
-------
125
125
116
100
c
.o
'
§ >
qj 0}
3 CL
LL. _
'5 o
o
0) =
O) »
CD O
0)
75
83%
60.5
50
43.5
25
21.1
1978
1979
1980
1981
1982
Figure II-7. Average specific fuel consumption for the Hartford operations, 1981-82.
Yotsuya Treatment Plant Case History. The six-hearth incinerator at this treatment plant
started operation in July 1979 and has since operated in an autogenous combustion mode. Sup-
plementary fuel is only needed to heat the furnace at the start of combustion. Incinerating capacity is
30 wet metric tons per day. Feed is dewatered on high-pressure BFPs. In 1981, 19,521 liters (5,157
gal) of heavy oil were used for 8,122 Mg (8,934 tons) of dewatered sludge at 30 percent solids con-
tent. This showed that, on the average, only about 8 liters (2 gal) of heavy oil were used to in-
cinerate one metric ton (1.1 tons) of dewatered sludge solids. This incinerator is clearly one of the
most energy-conserving in Japan, since the consumption of supplementary fuel (heavy oil) by in-
cinerators is commonly 170 liters (45 gal) per ton of solid matter for a MHF. The savings in heavy
oil achieved by autogenous combustion represent about 7 percent of the total operation and
maintenance cost for the treatment plant. Currently, the incinerator operates on a weekly cycle (i.e.,
it starts on Monday and stops on Sunday). Switching to long-term continuous operation is technically
possible, and, if this is done, the number of heatup times will be reduced and costs further lowered.
Discussing the operation in greater detail, thickened sludge at a concentration of about 4 percent
solids is dewatered by high-pressure BFPs after being conditioned with polymer at 0.5-0.8 parts by
weight per hundred (pph) and ferric chloride at 5-7 pph dosage rates. The solids content of the
11-27
-------
dewatered sludge is about 38 percent. Ferric chloride is added to improve the release of the
dewatered sludge from the filter fabric.
Dewatered sludge is fed from the sludge hopper into the top of the incinerator by a screw-type
conveyor. Both are closed structures. Closed screw-type sludge conveyors are reported to be ex-
cellent from the standpoint of odor control and appearance but have previously been considered un-
suitable for incinerators. They tend to form sludge balls that do not completely burn in the furnace.
This plant's designer reconstructed the rabble arm teeth so that the lumpy sludge is completely com-
busted to ash in the MHF. When the ash drops from the bottom hearth, it is moved by conveyor to
the ash hopper. The ash is removed from the plant by adding about 20 percent water by weight and
is disposed to a landfill.
Sludge combustion air is induced into the bottom stage of the furnace by draft. Additional air is
induced into the middle stage of the furnace. The exhaust gas is cooled and thus dehumidified,
scrubbed, and alkali-washed in a cyclone spray-type scrubber that uses secondary effluent and caustic
soda. It is then deodorized by passing it through acid and sodium hypochlorite scrubbers, further
scrubbed by a wet-type electrostatic precipitator, mixed with hot air from the shaft cooling of the
furnace to prevent a steam plume, and discharged from the stack.
Problems and Solutions
In early trials, incineration by autogenous combustion displayed a number of problems compared
with incineration using supplementary fuel. Even though the heat balance between combustible mat-
ter and moisture indicated that autogenous combustion was possible, stabilized autogenous combus-
tion operation could not be achieved in conventional furnaces. Three major problem areas included
unstable combustion, unburned sludge, and use of large quantities of air. The combustion
temperature and location shifted frequently due to the varying amounts of combustibles in the feed.
Since the combustion zone was narrow, due to a lack of heat from the burners, the sludge retention
time on the hearth was short, and the temperature was frequently low; therefore unburnt sludge
resulted. High air use resulted because the combustion temperature was controlled by the amount of
cooling air; as the air ratio increased, heat loss to the exhaust gas increased and thermal efficiency
deteriorated.
Stabilized autogenous combustion operation resulted from the introduction of the procedures
discussed here. The volume of primary air flowing into the bottom stage of the furnace can be
automatically controlled, taking advantage of the draft effect that changes according to load change
in the furnace and the change of sludge character. This widens the combustion zone and prevents the
lowering of the surface temperature and the discharge of unburned matter in conjunction with a
change in the combustion load. The hearth with the greatest temperature in the combustion zone can
be detected with a sensor and the volume of secondary air directly induced into the combustion zone
can be manipulated so that the temperature on this hottest hearth is held between 700° and 900 °C
(1,300° and 1,650°F). In times of low-load operation, the operation can be governed by moving the
combustion hearth up higher in the furnace. Thus, heat radiated from the furnace walls and the
center shaft can be reduced and the lowering of the thermal efficiency prevented. These control
measures prevent unstable combustion, high air ratios, and discharge of unburnt sludge; they also
minimize clinker and slag formations.
Performance Test of MHFs at Yotsuya. An experimental program was undertaken by the
Japanese to investigate the conditions required for autogenous combustion in a MHF. This was done
by changing both the amount and the solids content of dewatered sludge put into a furnace. The
sludge employed for the study had a volatile solids concentration of 70 percent, and its high heat
11-28
-------
value was 5,800 kcal/kg (10,453 Btu/lb) VS. The incinerator was designed to handle 30 metric
tons/day (33 tons) of dewatered sludge with a water content of 70 percent and a volatile solids con-
tent of 70 percent.
In the study, stabilized autogenous combustion with an input load range of 9.5-40.6 metric
tons/day (10.5-44.7 tons) was possible if the net heating value (NHV) of the input dewatered sludge
was 400 kcal/kg (720 Btu/lb) cake or more. This range was 32-135 percent of the design value;
thus, it was clearly possible to handle a very wide range of loads. In this test program the NHV of
the dewatered sludge was only 350-370 kcal/kg (631-667 Btu/lb) cake. Maintaining the condition of
autogenous combustion for more than 6 hours was possible in some cases but impossible in others.
This seems to show that if the sludge has a low calorific value of under 350-370 kcal/kg (631-667
Btu/lb) cake, autogenous combustion in the furnace is not practical (NHV takes into account the heat
sink effects of sludge moisture, combustion products, and excess air, all raised to furnace exit
temperature).
The properties of exhaust gas at the furnace outlet were found to be as follows:
NOX concentration 100-260 ppm, related to 12 percent oxygen concentration
HCN concentration ND-66 ppm, related to 12 percent oxygen concentration
Dust 0.2-1.4 g/Nm3
Odor concentration 4,100-7,300 in autogenous combustion; 17,000 in combustion using sup-
plementary fuel
It was found that odor concentration was lower in autogenous combustion than in combustion with
supplementary fuel. When the degree of conversion into NOX and SOX was calculated with respect to
N and S quantities in dewatered sludge put into the furnace, on the assumption that 10 ppm of ther-
mal NOX will be created at 800° - 850°C, the conversion from N into NOX was 2.0-4.7 percent, and
the conversion from S into SOX was 84-120 percent. These values generally agreed with NOX and
SOX production in existing furnaces. The unburned portion of incinerated ashes was nil to 0.67 per-
cent; thus the extent of sludge combustion was very satisfactory.
Hojin Treatment Plant Case History. At the time of its planning, the incinerator at the Hojin
Treatment Plant in Nagoya City was believed capable of autogenous combustion. However, when it
began operation in April 1979, it could not be operated in the autogenous mode. As the control
system for incineration was improved, however, oil consumption went from about 10 liters (2.6 gal)
per metric ton of dewatered sludge to less than 3 liters (0.8 gal), including that required for startup
or maintenance of temperature when not burning.
Thickened sludge with a solids concentration of 3-4.5 percent is dewatered by a horizontal
recessed plate pressure filter after dosing with ferric chloride at about 8 pounds per hundred pounds
(pph) dry sludge solids and slaked lime at 25-30 pph. The solids content of the dewatered sludge is
32-45 percent and the filtration yield is 3-6 kg/m2/hr of dewatered sludge solids. Because of its
relatively low water content, the sludge is mixed by a pug mill and fed to the incinerator. The
slablike sludge cake is crushed in the mill to permit easy handling on a conveyor. This improves
both drying and combustion in the furnace. The incinerator has 10 hearths, but the first does not
function as a drying hearth. It is provided for potential future use as an afterburning chamber for ex-
haust gas deodorization. Nine hearths are used to carry out the incineration process, and the fifth
and sixth are the usual combustion hearths. At present, exhaust gas is treated by water scrubbing,
alkali scrubbing, and electrostatic precipitation.
11-29
-------
Average incinerator operation conditions prior to upgrading were: furnace outlet temperature,
330 °C (625 °F); miscellaneous heat loss, 8.4 percent of the total heat input; sludge feed rate, 4.0
metric tons/hr (4.4 tons) of sludge solids concentration of 20-40 percent; and a calorific value within
the range of 2,500-4,000 kcal/kg (4,505-7,209 Btu/lb) dry sludge. Since dewatered sludge at the Ho-
jin Treatment Plant typically has a calorific value of about 2,600 kcal/kg (4,700 Btu/lb) dry solids,
the limiting dewatered sludge solids content in autogenous combustion was 34 percent, and the
limiting NHV was 450 kcal/wet kg solids (810 Btu/lb).
Upgrading of operations basically included the development of a combustion control strategy and
refurbishing to permit automatic control. The furnace temperature was formerly controlled using the
temperature in the third hearth as a reference. The fifth hearth, which was the combustion hearth,
was made the index. Temperature control in the sixth and seventh hearths also was made possible
because it was assumed that the combustion hearth could actually be the sixth or seventh hearth,
depending on the amount and properties of the dewatered sludge. The selection of the control hearth
is made manually. So that the gas temperature can be measured without being affected by flames,
the length of projection of a thermocouple into the hearth was reduced. Two thermocouples were in-
stalled in the fifth hearth, with a view towards higher measurement accuracy and ready detection of
the failure of either. Air for autogenous combustion was formerly supplied from the burner blower.
This was changed to a method that supplied combustion air in proportion to sludge input rate, by
positioning an air damper on a duct that branched from the central shaft cooling fan. The furnace
temperature can now be controlled by adjusting the volume of air handled by the exhaust gas fan
rather than by supplying tempering air from the burner blower. Instruments were added and their
control capability was improved. As a result, the variation in furnace temperatures has decreased,
and combustion has stabilized.
Autogenous Combustion Control System
Incineration is stabilized by controlling the air volume and furnace pressure. Three temperature
control levels, 800°, 850°, and 900°C (1,470°, 1,560°, 1,650°F), are available as temperature set-
points and are measured in the fifth hearth (or in the sixth or seventh hearths, if this is where com-
bustion occurs). The process control variables are air volume and furnace draft. Control is executed
by a closed temperature feedback loop: temperature control setpoint—process variable control—
temperature variation— process variable control. In the temperature range of 800° - 900 °C (1,470° -
1,650°F), which is the most desirable operating range, combustion air volume is controlled in pro-
portion to sludge input. Draft in the furnace for 800° - 850 °C (1,470° - 1,560°F) is fixed at -30
mmAq (negative pressure of 30 millimeters measured by a water column draft gauge). When the
temperature exceeds 850 °C (1,470°F), the furnace draft is gradually increased and cooling is ef-
fected. If 900 °C (1,650°F) is exceeded several minutes later, the proportional control of combustion
air is lifted and air for combustion plus cooling is supplied. If the temperature at the fifth stage
reaches 1,000°C (1,830°F), cooling air is increased by operating the burner blower. If 1,050°C
(1,920°F) is reached, sludge feed is stopped. In cases of temperatures below 800 °C (1,470°F), a
temperature rise is generated by reducing draft to -10 mmAq, which reduces the excess air and
quenching effect. If the temperature is still below 800 °C (1,470°F) several minutes later, combustion
air is manually reduced after lifting combustion air control in proportion to sludge input.
Properties of Exhaust Gas
The results of exhaust gas analysis at the outlet of the electrostatic precipitator are characterized
by a small emission of NOX and SOX. NOX emission was 0.2-0.4 g/kg of cake feed. SOX emission
was often below the detection limit value. At other sludge treatment plants in the city, where sludge
is dewatered by vacuum filters and incinerated by MHFs with auxiliary fuel, the NOX is typically
11-30
-------
0.6-0.8 g/kg of cake feed. As in the case of SOX, this seems to be due mainly to the fact that heavy
oil is not the main fuel.
SUPERAUTOGENOUS (HIGH-CALORIFIC) COMBUSTION
Burning sludge cake with a high net positive heat value requires special consideration. In the
past decade, a number of municipal sludge incinerators were built to combust sludge cake that has
more heat value than the water in the cake requires for evaporation and elevation of flue gas
temperature. These cakes are produced from either thermally conditioned sludge or raw primary
sludge, the latter perhaps containing grease and oil or admixed scum. The increasing adoption of
BFPs and recessed plate filter presses that can more effectively dewater sludges has generated con-
cern about burning this type of dewatered sludge cake in furnaces of the multiple-hearth type, which
were designed for cakes that require significant drying before combustion can begin. FBFs do not
raise these concerns, but have other factors that limit their acceptance.
These feeds are termed "hot" in the sludge processing industry. Other terms that apply are high
calorific and superautogenous, or simply autogenous meaning that the cake maintains itself in com-
bustion without the need for auxiliary heat sources. Sometimes burning these sludges produces a
high flue gas temperature at the furnace breech, which causes excess air emissions or higher than
desirable flue gas temperatures at the furnace outlet. Various approaches have been proposed by
combustion specialists in the sludge incineration field to manage the burning process within accep-
table limits, to prevent damage to the furnace, and to prevent episodes of massive combustion on
several hearths at once. All of these problems have in fact plagued some of the early projects.
Perhaps the first instance of serious damage to a furnace was the loss of the top two hearths at
Kalamazoo, MI, in 1974. This was caused by premature burning of the thermally conditioned sludge
in the topmost part of the furnace. The system had been put into service a few years earlier, and the
design was only suited for handling raw primary sludge cake; it was not appropriate for "hot" feed.
One of the first methods of control, based on the presumption that the volatiles needed more room
for combustion, was to feed to the second or even a lower hearth. This was done by providing an
enlarged drophole immediately below the feed entry chute or, in a few cases, by mounting a screw
conveyor in the sidewall of the furnace to inject cake laterally at the desired hearth. The screw con-
veyor approach, as applied in 1972 at Muskogee, OK, to handle thermally conditioned sludge filter
cake, suffered from burn-back difficulty and has seldom been used. Another project where a great
deal of difficulty occurred was in the MHF at Atlantic County, NJ. The sludge, after heat treatment,
had an unusually high heat value, and it was necessary to modify the furnace internally and provide
additional air for quenching.
A well-known furnace manufacturer has modified the design of the air input system to overcome
much of the past difficulty experienced with hot sludge. This design was applied as a retrofit at
Lansing, MI, and in a new furnace started up in late 1983 in Hawaii. Neither of these cases
represents a complete and perfect adoption of the design principles, but both have shown substan-
tially better performance than could be expected from previous furnace systems. A case history of
the retrofit at Lansing follows. It illustrates the performance of an incinerator that combusts high-
calorific sludge cakes.
Lansing, Ml, Case History
The original incinerator design at Lansing, MI, was for burning a nonautogenous sludge cake;
however, it could not satisfactorily operate within allowable particulate air emission standards while
burning the superautogenous sludge that was produced with the adoption of a thermal sludge condi-
tioning process. In 1978, the sludge was thermally conditioned and dewatered by a rotary vacuum
11-31
-------
drum filter to a 50-60 percent solids concentration. In addition, the heating value of dry cake was
7,200-7,800 kcal/kg (13,000-14,000 Btu/lb) compared to a 4,200 kcal/kg (7,500 Btu/lb) design value
normally experienced and used for sludge furnace design purposes in the United States.
Modifications to the furnace and its control system were completed in 1982 and corrected the
deficiencies. The main feature of this new design, known as Cyclo-Hearth® , is a distributed sludge
combustion air supply that allows air to be routed where needed for temperature management and
prevents overheating when burning a superautogenous cake. This air management system consists of
two elements: (1) individual supplies to each hearth that are readily modulated with accuracy by
automatic or operator intervention and (2) a swirling effect induced by air jets that operate when
burners are shut off to simulate the burner gas dynamics and thus promote the equalization of
temperatures laterally in a hearth. The Cyclo-Hearth MHF consists of a number of discrete furnaces
(or hearths) acting together in series rather than a single massive operation. The basic design of the
Cyclo-Hearth MHF (see Figure II-8) is implemented through the following provisions. Temperature
Auxiliary
Fuel
Burner
(Typ.)
High
Velocity
Mixing Jet
(Typ.)
*s;
Sludge
Combustion
Air
(Typ.)
Figure 11-8. Schematic of the Cyclo-Hearth® multiple-hearth furnace.
11-32
-------
is controlled on each hearth by modulating either the burner firing rate or the sludge combustion air
rate, but not both simultaneously. For instance, if a hearth is on burner control with the burners fir-
ing at a low rate and the hearth temperature rises above the set point, combustion control logic will
switch the hearth to combustion air control. Control logic will also switch in the opposite direction
as conditions warrant. Exhaust gas oxygen content (excess air rate) is maintained at the set point
value by allowing the oxygen analyzer to override the temperature control signal to the bottom
hearth. This will allow the bottom hearth sludge combustion air valve to modulate and admit more
air as required. High-velocity mixing jets impart turbulence to unfired hearths by directing a stream
of air tangent to a circle, dividing the hearth area approximately in half. Operation of the jets is only
required on unfired hearths and on fired hearths when an autogenous sludge is being incinerated. By
promoting turbulence, complete combustion, and uniform hearth temperature, these mixing jets allow
greater response to furnace conditions and eliminate control circuit instability caused by delayed
combustion. Furnace draft is controlled by varying the position of the adjustable throat in the Ven-
turi scrubber, to allow better particulate emission control with no increase in electrical power
consumption.
Modification work at Lansing, MI, was restricted to the furnace, sludge combustion air fan, air
distribution system, and sludge feed point. The waste heat boiler, scrubbing system, and induced
draft (ID) fan remained unchanged. The modified furnace is shown schematically in Figure II-9.
Three sludge combustion air ports were added to the first through third hearths, thus enabling this
air to be added wherever needed in the furnace. A damper was placed in each 25-cm (10-inch)
sludge combustion air line to the furnace. These dampers can be controlled by the temperature con-
troller on each hearth or manually by the operator. In the automatic control mode, each hearth
temperature would be maintained at its set point by the flow of sludge combustion air to that hearth.
To provide adequate sludge combustion air, a larger ID fan was needed.
Three mixing jets were added to the first through fifth hearths. The purpose of these jets is to
provide gas-phase turbulence in these hearths when no burners are being used. The mixing jet air is
taken from the burner combustion air lines. Mixing jet air is controlled manually with valves located
in each line to the furnace. The operators determine when and how much mixing jet air is to be add-
ed to the furnace based on the temperature profile of the furnace and visual observation of the hearth
conditions. No modifications were needed in the burner combustion air fan to accommodate this
change, since mixing jet and burner combustion air are not needed simultaneously. The sludge feed
point was moved to a lower hearth, allowing more gas residence time in the top hearths to achieve
adequate combustion of the volatiles. The sludge feed can now be varied between hearths three or
four as determined by the operator. The rabble arms were removed from hearths one and two, and
refractory feed chutes were installed to bring the feed cake from the top of the furnace to the new
feed hearths. A single burner was added to hearth one in case it was necessary to have an ignition
source to ensure combustion of volatiles remaining at that point.
Modification Results. Operation of the modified system began in March 1982, on a 5
day/week, 24 hr/day schedule. The system operated at a wet feed rate of 3.6-4.5 Mg/hr (4-5
tons/hr), with roughly a 50 percent total solids content cake, and was able to meet air emission re-
quirements. With the feed split between hearths three and four, there is adequate gas residence time
in hearths one and two to allow complete combustion of the gaseous volatile material. In addition,
the temperature of these "afterburner" hearths can be controlled by the supplemental air injection.
In a similar fashion, the feed hearth temperatures are also controlled by the addition of sludge com-
bustion air. The operators can control temperatures on these feed hearths at 540° - 650°C (1,000° -
1,200°F) even though the sludge is burning at these conditions. Maintenance of the desired
temperature profile in the furnace has reduced clinker formation to a minor operating problem.
11-33
-------
(
Sli
Com
Air
-*
1
*
1
]
1
i
— -* — y •*-
[3)
(3)
(3)
^ (3)
— »•
2?~
jdge
Dustion
Fan
L
D
SI
'l
Ash
ischa
udge Cak
Feed
=
e
rge
Center
Shaft
Air Fan
—*• / — »- To Main Stack
Hearth No.
1
2
3
4
5
6
7
8
(3) — ^- Breeching
3 (1 )
2 -
(3)
3
(3)
(3)
3
(3)
Mixing
Jets
(3)
"^i
(3)
I
Burners
i ^ ^^omi
> i
' Fuel
Burner
Combustion
Air Fan
Figure II-9. Modified Cyclo-Hearth® furnace configuration.
11-34
-------
The sludge feed distribution between hearths three and four is determined by the sludge cake
solids content. Should the temperature of hearth three increase in spite of large combustion air rates
to that hearth, indicating a hotter sludge, the operator diverts more sludge to hearth four, thus main-
taining the proper temperature profile. Conversely, when little combustion air is directed to hearth
three and that hearth temperature begins to drop, the sludge feed rate to that hearth is increased. The
modified furnace is also capable of handling fluctuations in feed quite successfully. If the addition of
an autogenous sludge is stopped for a brief period of time (e.g., 10-15 minutes), the operator simply
reduces the amount of air entering the upper portion of the system until the feed problem is cor-
rected. If the feed interruption persists, the volatile material in the furnace is depleted and the
temperature profile cannot be maintained by sludge combustion. The operator then actuates the
necessary burners to maintain the system's temperature profile. Meanwhile, the emission control
system is able to maintain the exhaust gas quality.
Control of the sludge combustion air to the furnace can be accomplished by two methods:
automatic controller and manual loading stations and manual valves. The operators currently run the
furnace in the manual mode. The automatic system, with the original instrumentation, tended to
overcompensate, which resulted in an oscillating temperature profile. Proper tuning of the instrumen-
tation eliminated this problem. The operators do not find temperature control difficult and are able to
keep the furnace operating in a stable mode. Another innovative feature of the modified furnace con-
cerns the method of sludge combustion air addition to the upper hearths. Sludge combustion air ports
on hearths one through three have been placed there for that purpose, while the mixing jets on
hearths one through five were installed only to promote good combustion for burning autogenous
sludge. Each 5-cm (2-in) mixing jet line operates at 0.1 kPa (16 oz/in2) header pressure and
discharges a significant amount of air into the hearth when the valve is open. Consequently, the
1.9-cm (0.75-in) mixing jets are used to provide gross temperature control on a hearth, and the
sludge combustion air manual loading stations are used for fire control.
The Lansing furnace was not equipped with sufficient instrumentation to verify the ability of the
mixing jets to equalize the temperature on a given hearth, but a visual examination did not indicate
problems of uneven temperature profiles with the modified system.
Cyclo-Hearth vs Conventional MHF at Lansing. The most significant improvement of the
Cyclo-Hearth over the conventional MHF is the degree of operator control over the furnace
temperature profile. When burning autogenous sludge, the conventional system was unable to process
the material satisfactorily due to inadequate hearth temperature control, which caused clinker forma-
tion in the furnace and resulted in frequent shutdowns for removal and high wear rates on the rabble
teeth. The modifications provide the needed temperature control at all times, thereby minimizing
clinker formation. The modifications also resulted in a system that demonstrates better stability
response to feed changes than the previous system. The modified furnace responds to changes in the
feed conditions to maintain stable operating temperatures. When processing the superautogenous
sludge using the conventional system, operators had considerable difficulty maintaining temperature
stability, since the furnace lacked the necessary hardware to effect this type of control. The improv-
ed stack emissions of the Cyclo-Hearth over the conventional MHF are the most important results of
the modifications. Due to poor temperature control and stability, the conventional system was not
able to meet the limit of 0.2 gm particulate/kg dry gas at 50 percent excess air. The performance
test for the modified furnace was well under this limit, since 0.032 gm/kg dry gas at 50 percent ex-
cess air was achieved.
WASTE HEAT RECOVERY EQUIPMENT
In early 1983, a new installation of four MHFs fitted with waste heat boilers began generating
steam at the Metropolitan (Metro) WWTP serving 57 municipalities in the major core of the Twin
11-35
-------
Cities area in Minnesota. The plant, located on the Mississippi River in Saint Paul, has a nominal
flow rating of 11 m3/s (250 mgd) and produces 81 Mg/day (90 tons) of raw primary sludge cake
conditioned with polymer and 131 Mg/day (145 tons) of thermally conditioned cake. Typical solids
content of the primary sludge after dewatering by twin-roll presses is 32-35 percent. Typical solids
content of the cake from the thermal conditioning and dewatering process, which handles a blended
sludge that is 3 parts secondary to 1 part primary, is also held at 32-35 percent. Process conditions
in the dewatering step that employs automated membrane presses can be set to achieve solids content
of 55 percent or more, but this burns too hot in the furnace and causes clinkering.
For the first 24 days in January 1984, steam production averaged between 14-18 Mg/hr
(30,000-40,000 Ib/hr), and total steam produced when valued at the equivalent cost of natural gas
that would otherwise be burned in onsite steam generators was worth $113,000. This is based on gas
costs of 0.4C/MJ ($4/MBtu). The 1984 goal is to produce steam valued in this way worth $1.5
million. If valued in No. 2 fuel oil terms at 0.26C/1 ($l/gal), and even allowing for a higher excess
air ratio to avoid smoke, this equivalent value would be approximately doubled. The system at the
Metro plant consists of two MHFs and two furnaces that had been in service for 13 and 9 years,
respectively, plus two 1968 remodeled standby furnaces not fitted with waste heat recovery. All are
designed to accommodate autogenous cake plus scum burning. The top hearth receives no feed and
is the "zero-hearth afterburner." All have provision for the addition of pressurized sludge combus-
tion air at hearths zero through four, seven, and eight. The waste heat recovery steam generators are
rated at 9 Mg/hr (20,000 Ib/hr) at 2.8 x 103 kPa (400 psig). Actually, they are delivering at about
half of that capacity because of reductions in furnace loading. All steam needed by this thermal con-
ditioning process is supplied by this recovery system, and much of the plant's winter building heat
load is also provided. In warmer months, steam will continue to be used for the thermal conditioning
process and air-conditioning; the remainder will be used in turbines that drive equipment such as ID
fans and boiler feedwater pumps.
Table II-4 lists the locations that have reported good-to-excellent results in the operation of
waste heat recovery equipment.
Table 11-4. Successful waste heat recovery installations.
Generating Steam over
125 psig
Generating Steam up to
125 psig or Hot Water
Heating Air for
Combustion Supply
Green Bay, Wl
San Mateo, CA
Cedar Rapids, IA
Davenport, IA
Dubuque, IA
Atlantic County, NJ
Honolulu, HI (Sand Island)
Lansing, Ml
St. Paul-Minneapolis, MN (Metro)
New Rochelle, NY
Hopewell, VA
Campbell-Kenton County, KY
Niles, Ml
Buffalo, NY
Erie County, NY (South
Downs)
Kansas City, KS (Plant
No. 20)
Louisville, KY
Ann Arbor, Ml
Duluth, MN
Redwood City, CA
Amherst, NY
Tonawanda, NY
Watertown, NY
kPa = 0.14465 psi
11-36
-------
CHAPTER III. COCOMBUSTION OF SLUDGE AND
SOLID WASTES
INTRODUCTION
The rationale for considering the joint combustion of sludge and solid wastes is that sludge, even
after dewatering by conventional methods, lacks sufficient heat value to balance the evaporative
burden of its remaining moisture when it is burned in a typical combustor. A further disadvantage is
usually the need to supply adequate heat to the combustion products and excess air so that odor
emissions are minimized. Solid wastes, on the other hand, typically generate more heat than is
necessary to burn them to innocuous products. By computation, a heat balance is reached when the
per capita quantities of solid wastes and sludge solids are combined, provided that the sludge is
dewatered to about 25-30 percent solids. This is an easy task if the sludge is raw primary sludge
alone, but the refined dewatering is much more costly if the sludge has been digested or if a compa-
nion amount of biological sludge is present. A further complication is that more sludge mass is
created when biological treatment is employed, and the overall cake from any dewatering process,
except that preceded by heat conditioning, is generally wetter. Thus, the technical feasibility of
cocombustion without the need for auxiliary fuel has to take into account the treatment plant process
train.
The trend toward cocombustion in those European countries where incineration is a well-
established and widely applied technology is a natural progression. Solid waste incineration is far
less accepted in the United States, and thus joint combustion technology is considered less. In part,
this is because many earlier solid waste incinerators were serious air polluters. In Minneapolis, MN,
for example, two municipal incinerators that were built in the 1930s under a Public Works Ad-
ministration program were commonly ignited each day by burning auto tires found in the collection
trucks. Public pressure from citizens who were bothered by the air pollution helped bring about a
landfill program that began in the 1960s and terminated incineration.
New York City, where a plethora of apartment house incinerators was installed, has historically
fought hard against municipal incinerators for refuse or solid wastes and, as a result, public accep-
tance of sludge incineration in the near future appears unlikely. A major report for the Interstate
Sanitation Commission (NY-NJ-CT) proposed sludge combustion but called it "pyrolysis," a hi-tech
name used incorrectly in this instance to mean starved-air combustion (SAC). Refuse was to be pro-
cessed into refuse-derived fuel (RDF) and blended in, as was done in the mid-1970s pilot work in
California for the Central Contra Costa Sanitary District. Neither the New York nor the California
work has proceeded to full-scale design for both technical and institutional reasons. In Europe, even
20 years ago, incinerators were built to a higher standard of quality, and air emissions were con-
trolled by costly methods. There, the ability to more economically process two waste products at a
single site with a single management staff and shared maintenance and support workers is seen as a
major advantage; cocombustion also allows a single operating permit to be issued by the regulatory
agency. This is much less likely to be realized in the United States, where it is common to have one
completely separate entity responsible for the sludge and another one for the solid wastes, with the
solid wastes frequently controlled by private ventures rather than public agencies.
Partial appeal of cocombustion rests on the assumption that fossil fuel costs for sludge solids in-
cineration are an insurmountable obstacle and that the heat value of the refuse is needed. However,
in recent years, better methods of dewatering sludge have become widely adopted. This was dis-
cussed and illustrated with case histories in the previous chapter. Furthermore, the concerns about
IIM
-------
cost and availability of fuel have dissipated, and, as was shown in the previous chapter, more effi-
cient incinerators for sludge can be designed and older ones can be upgraded. Thus, the technical
risks and institutional aspects of a joint combustion facility have taken on a greater significance.
Cocombustion of refuse and sludge solids in the United States is currently practiced in only three
locations using widely divergent technologies.
COMBUSTORS SUITABLE FOR CODISPOSAL
Most burning has been tried in mass-burn refuse incinerators of the usual configurations adopted
for refuse that has had only minimal preparation, such as removal of bulky objects and explosion
risks. Incinerators of the water-wall type that are suited for heat recovery are the more modern type,
although sludge could also be burned in a refractory-wall type incinerator. The action of the grate,
application of combustion air, and the means of dispersal of the sludge are critical elements.
The rotary combustor, a combination of a water-wall incinerator and rotary kiln, has been in-
stalled in only one location in the United States (Gallatin, TN) and is currently burning only refuse.
It should be excellent for handling a mixture of coarsely shredded, moderately prepared refuse and
sludge cake in that it has absolute purging of any noncombustibles. Thermal economy and stability
when burning sludge cake along with refuse would have to be proved. This U.S. installation has
been found to be too small to allow adequate burn time for raw refuse.
The modular refuse incinerator is generally much smaller than the typical mass-burn type, and
its practicality for codisposal has not yet been proven. However, a well-dewatered cake could be
proportioned into the solid wastes/refuse feed and be expected to burn satisfactorily in the available
residence time, provided that sludge cake lumps did not "case harden" and leave a core of wet
solids surrounded by crust or ash. Agitation of the bed in a modular unit is minimal, and sludge ball
discharge with the ash is possible.
Flash-dryer incinerator equipment has been used for many years to dry wastewater sludge, but
has been less successful in incineration. Because the process involves suspension burning of fine par-
ticles, such a system is not at all suited to mixtures of refuse and sludge. However, it can be used in
a design as the method of preparing sludge for mixing with refuse or for injection above the burning
refuse in a conventional incinerator. If the flash-drying method is considered, it can be made cost-
effective by using recycled incinerator gases as the heat source. If, for example, flue gases pass
through a high-pressure boiler and come out at 315° - 370 °C (600° - 700 °F), there should be suffi-
cient heat energy for an effective flash-drying operation.
The MHF, commonly applied in sludge combustion, is likely to be used if the wastewater
agency is in control of design or has such units in service already. In this situation, preparation of
the refuse needs to be more elaborate, usually requiring size reduction to 3-5 cm (1-2 in) maximum
dimension to minimize fouling of the rabble teeth, unless metals are effectively taken out by air
classification. The furnace is self-clearing of noncombustibles if the ash system has been designed
appropriately. European experience indicates that best performance is obtained if sludge cake is fed
to the top hearth and the RDF is fed to the normal combustion hearth. Flashback fires must be
guarded against in the RDF injection design.
The FBF was installed and operated as a cocombustion demonstration facility at both Duluth,
MN, and Franklin, OH. As noted previously, any design must deal with the noncombustibles that
especially tend to collect in a FBF rather than purge out automatically as in a traveling grate or
multiple-hearth unit. However, this problem can be materially lessened with proper design. At
Duluth, the problem of noncombustible matter collecting in the bed was aggravated by the design in-
adequacy of the classifier built into the secondary shredder. This caused a greater loading of non-
III-2
-------
combustibles than was expected. It was one of many materials handling aspects in the solid
wastes/refuse train that delayed the successful operation of codisposal and was not attributable to the
selection of a FBF. Similar problems at Franklin, OH, were solved within the first few years; then
the system ran well until shutdown (for other reasons) in the late 1970s.
Predrying the Sludge Solids
In the design of a cocombustion system, the thermal balance must account for the water present
in the sludge stream. If heat is to be recovered from the hot gases or if odor control considerations
require some preselected flue gas temperature, then the presence of more moisture than can be
tolerated will mandate that a drying step follow the dewatering of the sludge.
A further factor is the chosen means of combustion. Is it by suspension firing or burning on a
grate along with the refuse? Firing in suspension implies that the sludge solids enter the furnace
from the top, as a free-flowing granular or powdered product. This usually means that the residual
moisture has been brought down to the 5-25 percent range, which dictates the use of a direct drying
or multieffect evaporation step.
If grate firing is planned, a conventional dewatering step may be sufficient, as is employed at
Glen Cove, NY. It is necessary to get the sludge solids distributed evenly over the refuse charge. If
water reduction beyond that achievable by conventional dewatering is dictated by the process' ther-
mal balance, drying to 30-50 percent residual moisture may be sufficient. This can be accomplished
by indirect drying equipment, although such systems as applied in chemical, pharmaceutical, and
food industries have not been economically or technically attractive to the wastewater treatment in-
dustry for sludge use. Virtually all the sludge drying that is done in the United States, whether for
cocombustion (Stamford, CT; Flint River, GA) or marketing of a sludge product (Houston, TX;
Milwaukee, WI; Chicago, IL; Largo, FL), is by the direct drying method. A rotary dryer system of
two units was installed in 1982 at the Metropolitan WWTP serving the Saint Paul-Minneapolis area.
However, it is expected to be used only as a contingency mode of sludge disposal or if an
agricultural-use market is identified. At Harrisburg, PA, trials of an indirect drying device, termed a
hollow-flight jacketed dryer, have not given good results.
In short, drying of sludge makes cocombustion easier, but the need for the added capital invest-
ment and the incurring of operating and maintenance (O&M) costs have to be justified by the value
of the heat that is not needed for evaporation of water in the combustor. Energy is also required for
the heating of water vapor to flue gas temperature or, at a minimum, to the exit temperature of a
waste heat recovery device. Thus, the design decisions for a sludge drying process involve the same
reasoning as for applying a waste heat recovery system: the value of the heat and whether it can be
used beneficially.
U.S. COINCINERATION SITES
The following section describes the coincineration practices at several sites in the United States.
Case histories are presented for systems at Stamford, CT; Glen Cove, NY; Duluth, MN; Flint
River, GA; and some trial operation sites in the United States.
Stamford, CT
This coincineration system is the only operational one of its kind in the United States. It was
proposed in 1968, put into operation in December 1974, and has been operational ever since. Sludge
for this process is produced at the city's 0.9 m3/s (20-mgd) conventional activated sludge treatment
III-3
-------
plant. Using progressive cavity pumps, the mixture of primary and secondary sludges is pumped to
belt filter presses. Approximately 7.6 m (25 ft) upstream of the belt filter presses, poly electrolyte is
added into the sludge piping. This point of addition allows for better mixing of the sludge and
polymer. The conditioned sludge is dewatered to an average cake solids concentration of 26 percent.
The cake is discharged to a pug mill, where it is combined with previously dried sludge to produce
a mixture with a solids concentration of approximately 65 percent. This mixture is then conveyed to
a rotary dryer. A portion of the hot gas that would normally be wasted through the solid waste in-
cinerator stack is tempered with ambient air and introduced into the dryer at about the same location
as the sludge mixture. As the dryer rotates, the sludge is cascaded through the hot gases, and
moisture is evaporated at a rate of 2,300-3,200 kg (5,000-7,000 Ib) of water per hour. The dried
sludge, with a solids content of 90 percent, is discharged through a diverter gate and divided into
two streams, one of which is recycled to the pug mill while the other is conveyed to the incinerator
and burned. The heat value of the sludge averages 20,930 kJ/kg (9,000 Btu/lb) of volatile solids.
The incinerator is a conventional mass-burning refuse incinerator with rocking grates. It has a
capacity of 330 Mg/day (360 tons/day) and uses electrostatic precipitators for pollution control. The
refuse/solid waste as received with no pretreatment has a heat value as fed of 14,000 kJ/kg (6,000
Btu/lb). It enters the incinerator through a charging hopper and is discharged onto the grates at a
rate of approximately 180 kg/min (280 tons/day). At preset intervals, the grates rock, thus moving
the burning material through the furnace. Combustion is controlled using overfire and underfire air
fans. At the end of the furnace bed, the ash drops into a wet sluice and is conveyed to a truck for
landfill.
The dried sludge, at a rate of about 9.5 kg/min (15 dry tons/day), enters the furnace through
ports in the ceiling and burns in suspension within the first 1m (3 ft) of drop. The hot gases for the
drying system are drawn from the incinerator at 980 °C (1,800°F) at a rate of about 6.1 m3/s
(13,000 cfm) and are reduced to 200° - 400°C (400° - SOOT) by adding ambient air. This
temperature is controlled by the dryer exhaust temperature, which is set at 66-79.4°C (150° -
175°F). As the hot gases pass through the dryer, they pick up moisture and dust which must then be
removed in a cyclone dust collector. These gases are then returned to the furnace for deodorization.
Soon after this system went into operation, it became obvious that several modifications had to
be made to enable the system to work effectively. Many of these were quite simple and most were
designed and installed by plant personnel. The coincineration system was installed in the existing in-
cinerator building. Because space was limited, it was necessary to spread the equipment over five
different floor levels. Many conveyors were needed to move the sludge from one stage of the pro-
cess to another, thereby increasing material handling problems.
At various stages in the system, samples were taken to determine material moisture. It was
observed that as material moisture increased, the amperage of the conveyor motors increased. In ad-
dition, if the operator tried to process too much sludge, the amperage would also increase. Using
these facts, an amperage range was established for ideal sludge moisture content and volume. Por-
table ammeters were used for the determination. After operating for several days to prove that these
ranges were correct, permanent ammeters were installed at the main control panel to monitor the
pug mill and all critical conveyors. These allow the operator to control the process from the main
panel and determine whether to increase or decrease the dry recycle and dewatered cake rates on the
basis of the amperage reading. This has reduced operator fatigue and allows for a more stable pro-
cess. The operator is still required visually to inspect the entire system but at much less frequent
intervals.
Clogged conveyors presented another serious material handling problem. This was caused by rag
buildup on the bearings, changes in the material characteristics, and broken drive belts. As the con-
III-4
-------
veyor began to get clogged or if the drive belts broke, the rotational speed of the shaft would
decrease or completely stop. Again, this was fairly simple to solve. By attaching speed sensors to
the drive shafts of the gear reducers, this decrease of speed could be sensed. Now, as soon as the
speed decreases below a certain point, the electrical interlock system shuts down the material feed to
that conveyor and an alarm rings at the main control panel, alerting the operator and allowing him
to take action before any serious equipment blockages occur. Because of this, downtime has been
greatly reduced, and the operator is no longer faced with the frustration of having to remove com-
pacted sludge from the conveyor.
Another material handling problem was caused by the thixotropic nature of dewatered sludge
cake; that is, the sludge would change in viscosity as stress was applied. A screw conveyor approx-
imately 20 m (65 ft) long was used to convey the dewatered sludge from the presses to the pug mill.
Problems were encountered almost immediately. The physical characteristics of the sludge cake
began to change dramatically as the sludge proceeded through the conveyor. The sludge became very
sticky, making it difficult to convey and causing excessive torque on the drive motor. It was virtual-
ly impossible to run the system. The screw conveyor had to be replaced by a belt conveyor that
would not alter the physical characteristics of the sludge. Short screw conveyors for dewatered cake
do not seem to affect the sludge characteristics, but as the length of the conveyor increases, this pro-
blem becomes more evident.
Dewatered cake dryness and polymer concentration in the cake also appear to have considerable
effects on the ability of the system to function. Instead of the usual type of dried material, which is
light and fluffy somewhat like the material inside a vacuum cleaner, the sludge begins to form balls,
initially about the size of peas. These balls are dry on the outside and moist on the inside. As these
balls circulate through the drying system, they get larger and larger. Surface area is reduced con-
siderably, resulting in a greater recycle ratio by weight. This, at times, also makes the system vir-
tually impossible to operate. High polymer dosages also tend to make the sludge sticky, creating
drag on the conveyors and difficulty in mixing in the pug mill. Experiments have shown that these
changes occur as the cake solids concentrations drop below 22 percent or the polymer dosage in-
creases beyond 10 gm/kg (20 Ib) of dry polymer per kg (ton) of dry sludge. Therefore, when
designing this type of system it is important that dewatering equipment is specified that will obtain
the desired cake solids. Care in selecting and evaluating polymers will ensure dosage below this
level.
Fires were another serious problem. Most of the fires occurred inside the dryer. To control
these, an automatic water spray system was devised that includes a spray bar located across the
diameter of the feed end of the dryer and two sprays located at the discharge end of the dryer. It
was critical that no water from the spray system be allowed to impinge on the periphery of the
dryer, since thermal shock could possibly cause damage. Five stainless steel fogging nozzles are
spaced evenly along the spray bar. Three nozzles are adjusted to spray the length of the dryer and
two are adjusted downward. The sprays are controlled by a thermocouple in the exhaust end of the
dryer. When the temperature exceeds the set point of 150°C (300°F), a solenoid opens, allowing
water to flow through the sprays, which are set to pulse in intervals of 10 seconds on and 5 seconds
off or can be run continuously in a manual mode. The nozzles at the discharge end are also con-
nected to this system. The combination of the sprays has effectively controlled most fires.
Additional minor problems included spalling of metal from the dryer riding rings and hot spots
in the live-bottom storage bin. The problem of spalling was corrected by the installation of graphite
blocks on each of the four trunnion rolls. This small amount of continuous lubrication has prevented
serious wear on the rings. The hot spots in the corners of the live-bottom bin, which were a source
of smoldering sludge, were corrected by welding a sheet of metal inside the bin to round the corners
preventing a buildup of hot recycled sludge.
III-5
-------
In conclusion, although many operational problems have occurred with this system, most of
them have been solved, and the system now runs effectively and efficiently. When the plant was
built, the coincineration system was the only means of sludge disposal. An alternative method was
added to provide backup for the coincineration system. This alternative method is a postdewatering
lime stabilization system that costs approximately $1 million/yr to operate, mostly due to the cost of
trucking the sludge 65 km (40 miles) to the only available landfill site. The coincineration system is
extremely economical and energy conservative, since neither the incinerator nor the dryer requires
any external fuel source. It has not added any additional ash handling problems, and it is not a
source of air pollution. Additionally, the dried sludge before incineration can be used as a soil con-
ditioner where a market is available, thus creating a source of revenue for the municipality. Further-
more, excess waste heat can be used to generate steam or electricity, which could supply the
municipality-owned buildings, with the excess being sold to the public utility. The experience at
Stamford indicates that coincineration can be a viable means of sludge disposal.
Glen Cove, NY
One of the newest installations in the United States for joint combustion of refuse and sludge
solids is at Glen Cove, NY. This design is different from those at Stamford, CT, Flint River, GA,
and Duluth, MN, in that here the sludge is dewatered by centrifuge and fed in semiliquid form at
15-25 percent solids directly onto the refuse charge. The intent of the design is that it will evenly
distribute the sludge as a layer on top of the charge, with the sludge having a fuel value that is fair-
ly proportional to that of the refuse on a unit area basis. The Glen Cove plant has two mass-burning
furnaces, each designed for a daily feed of 100 Mg (110 tons) of refuse and 14 Mg (15 tons) of cen-
trifuge cake at 20 percent solids, or 3 Mg (3 tons) of dry sludge solids per day. If the refuse is
assumed to have 28 percent moisture, the dry weight ratio is 24 to 1 as compared to a wet weight
ratio for both of 7.3 to 1. If the refuse, as received, is compared to dry sludge solids, the ratio
would be 37 to 1.
Because the Glen Cove WWTP is not receiving flow as high as its design rate, the amount of
sludge to be disposed of is much less than the design basis. Actually, only about 2.7-7.2 Mg/d (3-8
tons/day) of 18-20 percent solids centrifuge cake are being fed to the furnaces, which are burning
refuse at their design rate total of 200 Mg/d (220 tons/day). Thus, the charging ratio is higher than
design, so the sludge is being handled readily. As a result, the operation of this system is not
presently representative of the per capita ratio of about 14 Ib of dry refuse to 1 Ib of dry sludge, ex-
pressed on the "as received refuse to sludge dry solids" basis. Instead, there is so much excess
refuse compared to sludge that the ratio is greater than 50 to 1. This does not represent a fair
demonstration of the adequacy of this design to handle a balanced per capita basis feed.
The Glen Cove furnaces are the refractory-lined, mass-burning type, equipped with "Kascade"
stokers and automatic combustion air controls to maintain uniform combustion temperatures and con-
ditions. Each furnace is equipped with a 9,080 kg/hr (20,000 Ib/hr) convection boiler rated for 41
kg/cm2 (600 psi) at 250°C (480°F). Available steam is converted to electric power in a 2,500 kW
multistage condensing turbine generator set, which powers the complex of incineration facility and
the adjacent wastewater plant. Excess power can be sold to the Long Island Lighting Company. The
design basis sludge/refuse mixture has an average high heat value of 9,500 kJ/kg (4,100 Btu/lb).
This presumes 11 Mg (12 tons) of water and 3 Mg (3 tons) of sludge solids in the sludge stream to
each furnace. Of course, if the sludge is wetter than the 20 percent level, this heat value would be
reduced. At 16.7 percent, for example, there would be an extra 3 Mg (3 tons) of water to be
evaporated, which would reduce the heat recovery potential and thus the yield of electric power. At
15 percent solids, there would be an extra 4.5 Mg (5 tons) of water to evaporate. The incentive,
then, is to maximize the solids in the centrifuge cake consistent with getting even distribution over
III-6
-------
the refuse charge. Sludge cake above the 20 percent solids level is only achieved when straight
primary sludge is centrifuged.
Air pollution control is for particulates only, using two field electrostatic precipitators, each siz-
ed for 14,160 L/s (30,000 scfm) at 316°C (600°F). The performance guarantee is to achieve 0.11
g/m3 (0.05 grains/scf) at 12 percent carbon dioxide. The long-term performance of this particulate
removal system will be followed with interest by professionals in the wastewater field, where wet
methods of scrubbing are universally applied to sludge combustion gases.
Both the coincinerator system and the wastewater plant are operated for Glen Cove under con-
tract by a private company. The city estimates a savings of $700,000 per year compared to prior
costs for wastewater and refuse management. The project cost was $24 million.
Fluidized-Bed Coincinerator at Duluth, MN
Because of the fuel oil shortages resulting from the oil embargo of 1974, the Western Lake
Superior Sanitary District in Duluth, MN, and the district's consulting engineers decided to utilize
available solid waste as auxiliary fuel to incinerate sludge produced at the wastewater treatment
facilities. The construction of the facilities was completed in 1979, and the startup began in
November 1979. Performance tests were conducted in June 1980.
The Duluth installation consists of two systems, each having a CWB-type conventional fluidized -
bed reactor with a 6-m (20-ft) diameter bed and 10-m (34-ft) diameter freeboard, dual gas cyclones,
waste heat boiler, Venturi scrubber gas cooler, fluidized air blower and induced draft (ID) fans, and
heat exchanger for plume suppression (Figure III-l). The fluidized air blower, ID fan, and some
pumps are driven by steam turbines powered by the steam from waste heat boilers. The waste heat
boilers are rated for 21,792 kg (48,000 Ib) of steam at 19 kg/cm2 (280 psig). The boilers are the
two-drum, water-tube, natural circulation type, designed for 870 °C (1,600°F) gas inlet. The units
are equipped with soot-blowing capabilities. The sludge was to be fed into the fluidized-bed reactor
through a feed chute at the top of the reactor, and the refuse derived fuel (RDF) was to be fed with
a pneumatic solid waste feed system for injection into the freeboard approximately 0.6 m (2 ft)
above the fluidized bed, pointing downward to the bed. The gas-cleaning equipment consists of dual
gas cyclones, a Venturi scrubber, and a collector/cooler with three impingement plates. The system
includes a hydraulic ash handling system consisting of an ash slurry tank, an ash classifier, and a
fine ash thickener. The RDF preparation system includes coarse and fine shredders, air classifier,
magnetic separators, conveying and metering equipment, and a storage silo.
After completing construction and making the system operational, various attempts were made to
incinerate RDF without the sludge. Troubles developed in the RDF feed system, and undesirable
"freeboard burning" was experienced (i.e., combustion occurred too high in the furnace). As a
result of a malfunctioning air .classification system, large metal or wood objects constantly jammed
the rotary air lock at the RDF pneumatic feed system. The secondary shredder was designed to
shred 95 percent of the material to a size of less than 3.8 cm (1.5 in) in effective diameter. The
solid waste processing facilities had no presorting capability. Therefore, glass and other noncom-
bustible objects entered the primary shredder and were broken up and imbedded in the RDF. The
location of the RDF feed nozzle was incorrect. The majority of RDF burned in the freeboard area,
while fuel oil had to be burned to maintain reasonable bed temperatures. The addition of sludge, by
dropping sludge cake into the bed, made the situation worse. While trying to keep the bed hot by
burning fuel oil, the startup operators were also trying to keep the freeboard cool by spraying a fine
spray of water in the reactor. Fortunately, there was an alternate RDF feed nozzle located near the
top of the fluidized bed and directed downward to the bed. The point of entrance into the reactor
III-7
-------
03
03
-C
03
O3
03
c
03
03
•4-J
U3
05
cc
Q
CC
03
O3
CO
2?
5
D)
\ /
III-8
-------
was approximately 0.6 m (2 ft) below the top of the bed when the fluidized-bed height is 2 m (6 ft).
Although use of this optional feed location made considerable improvement, the freeboard
temperatures were not controllable when coincinerating sludge with RDF. Also, the necessary bed
temperature for combustion could not be sustained when coincinerating.
It became apparent that using the single RDF feed port at the lower location did not satisfy the
original design intent of coincinerating RDF and sludge cake in the bed. Multiple feed points for
RDF needed to be developed. The first approach was to modify the sludge feed port and utilize the
moisture content of sludge to control the freeboard temperature in lieu of freeboard sprays. A series
of attempts were made to spread the sludge over the freeboard. Finally, a special nozzle was
developed to feed the sludge in a small enough particle size to allow the water to evaporate and a
large enough size to have solids end up in the bed to complete incineration. The feed device
developed produces sludge particles approximately 0.3 cm (1/8 in) in diameter most of the time.
Steam is utilized to shear the sludge after it is extruded through a converging nozzle. The extruded
flow is not even, and periodic surges of dispersed sludge take place. With the new sludge feed
device and the RDF feed nozzle, experiments were made with the height of the bed. The bed was
allowed to grow as deep as permitted by the capacity of the fluidizing air blower.
Performance tests successfully demonstrated that the system could operate at design rates while
meeting particulate emission requirements. They also showed that the system was capable of in-
cinerating sludge with fuel oil, sludge with RDF, RDF alone, or all fuels at one time and at adjusted
loading rates. Table III-l shows the average quantities of RDF sludge, and fuel incinerated during
tests. The design capacity tests were conducted for an 8-hour duration. Ash collected during the test
was analyzed in detail and results are shown in Table III-2. Results of emission tests and exhaust gas
composition are shown in Table III-3. While coincinerating sludge with RDF, the excess air rate was
between 48 and 51 percent. The excess air rate during incinerating sludge with fuel oil was about 85
percent. While burning RDF without sludge, fluidizing and cooling requirements resulted in high
levels of excess air. The amount of excess air used during the RDF-only mode was 174 to 232 per-
cent. The visible emissions from the stack were limited to 20 percent opacity. During the tests, 98
percent of the time the opacity was within 10 percent, and only during the last 7 minutes of the
sludge/RDF test did the opacity exceed the 20 percent limit.
During the startup and testing period, the ash handling system was very difficult to operate. The
fine ash thickener continually plugged up, and a small wet cyclone eventually had to be added to the
system to relieve the problem. The fine ash refused to settle in the ash classifier, and the cyclone
dip legs had a tendency to plug near the ash quench tank. In addition, the inert fraction of the RDF
was too big to be elutriated, and the bed continued to grow in size. Arrangements were made to
remove some of the bed regularly, which further aggravated the ash system operation.
During the startup and test burning RDF-only mode, one of the gas cyclones plugged with slag.
The material looked like lava and was named "moonrock" by the plant personnel. It was at least
0.9 m (3 ft) high and located at the bottom of the cyclone. The sample of slag was very hard, black
in color, shiny, and amorphous.
It appeared that the slag material contained some amount of alkali metal silicates, which became
sticky viscous glass when heated. Although there were considerable amounts of A12O3, CaO, and
Fe2O3 present to react with low-melting, alkali metal silicates like Na2O»3SiO2, there were not
enough quantities to convert all of the metal silicates to metal oxide-alkali oxide-silica dioxides. The
retention time in the bed was not long enough to complete the reactions, and burning was taking
place in the ducts and cyclones. The residual amount of low-melting, alkali-metal silicates caused the
ash to agglomerate in the cyclones.
III-9
-------
Table 111-1. Average quantities of refuse-derived fuel, sludge, and fuel incinerated
during the performance tests.
Refused-derived fuel
Sludge
Refused-derived fuel
Sludge
No. 2 fuel oil (gal)
Sludge
Refused-derived fuel
Note: Ib x 0.454 = kg
Btu/lb x 2.3255 =
Sample
Elements2
Si
Al
Fe
Ca
Mg
Na
K
Ti
Wet
Ibs/hr
17,700
27,217
5,285
15,778
338
29,880
13,787
Id /kg
Table III-2.
Bed
Material
4/25/80
P.C.5
6.0
2.5
7.5
1.0
2.5
0.5
0.4
Dry Percent
Ibs/hr Solids
12,938 76.66
5,852 21.50
3,932 74.22
3,281 20.34
6,884 23.00
10,453 76.10
Ash and bed analysis'1.
Bed
Material
4/26/80
P.C.5
3.5
3.5
7.5
0.85
2.25
0.5
0.7
Percent
Volatiles
68.83
56.30
69.47
63.62
58.80
67.20
(Percent)
Cyclone3
Deposit
4/28/80
P.C.5
17.5
5.0
10.0
1.5
4.0
3.5
1.25
Heating Value
Btu/lb
8,652
10,468
8,963
10,735
19,400
9,884
9,464
RDF
Ash4
4/25/80
P.C.5
8.5
5.0
8.5
1.25
5.0
4.0
0.85
1 During RDF incineration without sludge; semi-quantitative spectrographic analysis
2 As Oxides
3 Melting point of material is 1,000°C (> 1,832°F)
4 Residue after ignition @ 590 °C (1,100°F) is 22 percent w/w
5 Principal constituent
III-10
-------
Table 111-3. Results of emission tests and exhaust gas composition.
FILTERABLE PARTICULATE CONCENTRATIONS AND EMISSION RATES
Concentration Emission Rate
GR/DSCF Ib/ton dry feed
Sludge with refuse derived fuel 0.028 0.468
Sludge with fuel oil 0.024 0.901
Refuse derived fuel 0.009 0.234
Note: 1 grain = 0.065 grams
1 cubic foot = 28.32 liters
1 Ib/ton = 0.5 gm/kg
EXHAUST GAS COMPOSITION* (Percent Volume)
Sludge with refuse
derived fuel
Sludge with fuel
oil
Refuse derived
CO2
11.5
7.9
6.0
02
7.2
10.0
14.2
CO
< 0.1
< 0.1
< 0.1
Balance
81.3
82.1
79.8
Moisture
26.55
12.1
9.9
*At scrubber exhaust, dry basis except percent v/v moisture
At the time the tests were conducted, FeCl3 and lime were being used to condition the sludge
for vacuum filtration. It was known that the addition of both Fe and Ca would help to further con-
vert the remaining alkali metal silicates. CaO with sodium silicate will form Devitrite
(Na2O-3/CaO6SiO2), which melts at 1,030° C (1,886° F); Fe2O3 reacts with sodium silicate to
form Acmite (Na2O'Fe2O3«4/SiO2), which melts at 955°C (1,751°F). With this assumption, it was
decided to continue the coimcineration tests after the slag was removed from the cyclone. The addi-
tion of limestone and clay to the feed to prevent this scaling gave encouraging results. Similar
results were reported using clay as an additive when incinerating sludges high in sodium. This
eliminated the buildup of molten salts on the bed particles and the resultant bed stickiness. During
the sludge/RDF coincineration, no slagging problems were encountered. Immediately after the com-
pletion of the performance tests in July 1980, the solid disposal facility was shut down for modifica-
tions to the RDF preparation and the ash handling systems.
In conclusion, the startup and modification work at Duluth showed that processed refuse and
wastewater sludge cake could be coincinerated and produce recoverable energy while meeting permit
standards for air emissions. In early 1984, the district purchased wood chips and bark waste from
III-11
-------
forest product plants in the vicinity and began coincinerating this material with sludge cake. New
belt presses have been started up successfully, replacing the vacuum filters in service, and the
resulting cake is higher in the ratio of volatile solids to moisture. Presently, the cost of the added
fuel in the form of wood industry waste is about $10/ton of wet cake at 16-18 percent solids. This is
considered less expensive than the cost of processing refuse in the RDF-making facilities.
Coincineration at Flint River, GA
The incineration of sludge solids and green wood chips at the Flint River WWTP in Clayton
County, GA, differs from the other U.S. projects described in that the sludge is dried and pelletized
before being fed to the combustor. The hot gases from the Pulse Hearth™ furnace are used to dry
the sludge in rotary triple-pass Heil dryers. The furnace system was started up in late 1982; the
dryers were existing equipment that had previously been heated by natural gas. The rated capacity of
the system is 25,000 MJ/hr (24 million Btu/hr) input. Two furnaces are arranged in series and pro-
duce a 980 °C (1,800°F) off-gas temperature that is tempered down to 540 °C (1,000°F) by mixing
with ambient air before entering the dryers. Exit gases are wet-scrubbed with water only, at a
pressure drop of 15 inches water column, and the air emissions limit of the plant is met. No signifi-
cant odor complaints have been reported. The manufacturer of the furnace claims that the multipass
design of the combustion chamber allows controlled burning of the sludge pellets and wood chips
without creating submicron fines or permitting unburned hydrocarbons to escape. Selection of a
cocombustion system like this would be most attractive in an area where forest products are
manufactured and where a steady supply of the wood chips at an attractive cost can be assured. This
is the case at the Duluth installation.
TYPES OF CODISPOSAL SYSTEMS IN THE UNITED STATES AND EUROPE
Although there are a number of different approaches to codisposal by incineration, most of them
can be categorized into four basic types: Direct Drying - Suspension Firing (DD-SF); Indirect Dry-
ing - Suspension Firing (ID-SF); Direct Drying - Grate Firing (DD-GF); and Indirect Drying - Grate
Firing (ID-GF). Each of the four types involves predrying the sludge cake before feeding it to the
combustion chamber. A fifth type, that does not involve sludge cake predrying, is exemplified by the
Glen Cove, NY, and Duluth, MN, projects in the United States but is not readily evident in Euro-
pean practice.
Figure III-2 illustrates the DD-SF type, where sludge coming from a wastewater treatment plant
(WWTP) is fed into a thermal dryer prior to incineration. In some cases this occurs directly,
whereas in other cases preliminary dewatering is accomplished by mechanical means before thermal
drying. Often a conditioning agent, such as a poly electrolyte (polymer), is added to facilitate
mechanical dewatering. Because these agents are of an organic nature, they do not adversely affect
subsequent incineration. Inorganic agents such as lime and ferric and aluminum salts would be
detrimental to heat value but may prevent rapid slag formation and be beneficial in that way.
However, chloride content is very undesirable. In the dryer, the solids concentration of the sludge is
raised typically to the 75-95 percent level. To promote drying by assuring proper consistency, the
feed sludge is often ground and mixed with a recirculated portion that is already dry. Finally, the
sludge powder is injected or blown into the incinerator in a zone above the burning refuse. Most of
the sludge powder burns in suspension, and complete destruction of all putrescible matter is virtually
assured. The energy required for drying the sludge is provided by removing a portion of hot flue
gases that are then directly brought into contact with the wet sludge. The quantity of flue gases must
be sufficient to furnish all the latent heat of vaporization required by the sludge to be dried. This
heat must be dissipated by evaporating water before the sludge gets dry enough to catch fire or
explode.
Ill-12
-------
Q
Q
CD
O
OT
O
+J
2
0)
c
'o
c
"ro
(0
o
a.
o
O
oi
0)
D
O5
*- flj
(/) fli
CO 2.
III-13
-------
During thermal drying, some of the sludge solids volatilize in the form of gases and vapors.
These highly odoriferous substances are returned to the hot zone of the incinerator, usually at a
temperature of 774°C (1,425°F) or more, so that they are safely destroyed by thermal means. If the
highest degree of environmental control is required, as in West Germany, an auxiliary firing system
is installed. A temperature control system monitors the final combustion temperature inside the in-
cinerator. In case this temperature drops below the minimum of what is considered necessary for the
destruction of odoriferous substances (i.e., 802°C [1,475°F]), auxiliary gas- or oil-fired burners will
light up to compensate for any energy deficiency. Thereafter, the mixture of flue gases coming from
both refuse and sludge burning is conveyed through a common air pollution control (APC) system
out through a common stack. The resultant ash contains the inert constituents of both the refuse and
the sludge. If a high degree of thermal destruction is achieved, the amount of carbonaceous matter
will be below 3 percent by weight, and the amount of putrescible matter will be below 0.3 percent
by weight.
There are two major options with regard to the DD-SF type. One is to recover additional energy
for steam generation. This is achieved by inserting a waste heat recovery boiler between the in-
cinerator furnace and the APC system. This has been done successfully and on a large scale in
Europe. The other option is to siphon off either all or some of the dry sludge powder and use it as a
soil conditioner for agricultural and construction purposes.
The ID-SF type differs from the DD-SF type in one important respect. There is no direct con-
tact between the hot flue gases from the incinerator and the wet sludge to be dried. This is
accomplished by means of a heat transfer fluid that is used to extract sufficient heat from the in-
cinerator's hot flue gases. This heat is then transferred to the dryer, where it provides the latent heat
of vaporization. The heat recovery equipment is installed on the outlet side of the incinerator. It may
include steam generators, hot water boilers, and heat exchangers. Steam, water, and oil are com-
monly used as the thermal transfer fluids. As before, the gases and vapors that emanate from the
dryer are returned to the incinerator for thermal destruction. The dried sludge is also conveyed to
the incinerator for suspension firing. The main advantage of the ID-SF type is that sludge drying and
sludge incineration are physically separated. This is particularly helpful in retrofit situations, where
space limitations require separate equipment installations.
A third type, DD-GF, is considered by many engineers as the simplest approach to sludge
disposal in a municipal refuse-fired incinerator. Sludge is received from the treatment plant and par-
tially dewatered by the aforementioned means. The resultant sludge cake is simply transferred to the
incinerator and burned together with the refuse on a grate. The precise mechanism by which this
sludge is added differs from one plant to the next. In one case sludge is mixed with refuse in the
pit, while in another case sludge is dropped into the feed chute above the refuse. Attempts have also
been made to simply spray this sludge into the incinerator at various locations. In all cases, the heat
needed for vaporization of moisture must be transferred to the sludge cake, or more precisely, to
each sludge particle, by direct contact. The effectiveness of any particular approach has aroused con-
troversy among a number of investigators. This is mostly because of the unique heat and mass
transfer phenomena which govern sludge drying.
The ID-GF type is similar to the combustor design discussed above, except that, in addition to
mechanical dewatering, the sludge is dried by indirect thermal means before it is fed to the in-
cinerator. The ID-GF type lends itself to retrofit applications. Here, heat energy is extracted from
the incinerator and transferred to the thermal dryer, which is similar to the process described for the
ID-SF type, except that very fine particles are not needed when grate fired and more residual
moisture can be tolerated. Indirect drying takes place in a separate device under conditions that can
be more closely controlled. Many investigators have established the critical importance of the three
T's (temperature, time, and turbulence) in effective sludge processing. Depending on the design
III-14
-------
features of the dryer, intimate mixing and agitation of individual sludge particles are promoted for
maximum thermal efficiency.
U.S. AND EUROPEAN CODISPOSAL INCINERATOR SITES
A number of experiences in burning sludge solids and refuse have been reported over the years.
Reasons for the discontinuation of coburning or failures at some of the reporting sites vary; in some
cases it was because not enough effort and investment went into the planning and operational stages.
Tables III-4 through III-6 are a comprehensive listing of these trials in the United States, and Tables
III-7 through III-10 are a listing of those in Europe. Data are presented for the four basic types of
codisposal incinerators discussed previously, with the exception that no U.S. incinerators of the ID-
SF type were identified.
Case Histories of European Installations
The descriptions of the European cocombustion facilities in this section expand on the tabulated
data in the previous section (Tables III-7 through III-10) and should assist in their interpretation.
European cities have adopted the principle of cocombustion much more than have U.S. cities. The
motivations are intense in Europe, and political jurisdictional problems have not been an obstacle. As
a result, solutions have been found that are grounded in technical and economic imperatives. Also, a
high quality of construction has been observed by visitors to European installations. The case
histories that follow present experiences of cocombustion installations at Bielefeld, Goppingen, and
Marktoberdorf in the Federal Republic of Germany and Dordrecht in the Netherlands. A com-
parative evaluation is then presented of seven additional European mass-burning facilities.
The Bielefeld refuse incinerator consists of three units with a capacity of 16 tons/hr each at a
calorific value of 11,000 kJ/kg. With respect to incineration itself, it is a rather conventional plant,
consisting of a bunker, the incinerators, boilers, electrostatic precipitators, wet scrubbers, and a
specially designed wastewater plant. The four following factors convinced the Federal Government to
permit the development: waste was pretreated in order to homogenize it and to mix it with sewage
sludge; scrap was separated prior to incineration; hospital waste was incinerated with other wastes;
and the wastewater treatment system included flocculation and heavy metal ion exchange processes.
The coincineration process design was based on the following throughput:
Household waste 195,000 tons/yr
Commercial waste 85,000 tons/yr
Sewage sludge (40 percent dry matter) 30,000 tons/yr
Hospital waste 10,000 tons/yr
Tons X 0.90718 = metric tons
The investment costs (1983) were 146 million Deutsche Marke (DM) ($58.4 million), about 30 per-
cent of which apply to environmental protection devices. The innovation of interest here is the
pretreatment of the waste. Combustion, energy generation, heat exchanger burden, and flue gas
cleaning are better equalized than in the incineration of untreated refuse. During the mechanical
pretreatment of waste, predried and digested sludge is added in such a way that the mixture could be
incinerated on conventional grate systems without difficulties. Scrap is separated before incineration
to gain better scrap and less heavy metal emissions in the raw flue gas. Raw waste and the sludge
(about 10-15 percent sludge by weight as wet cake containing 40 percent dry matter) are fed into a
Losche ball mill, which has an inner diameter of 6.5 m and is filled with 50 tons of 12-cm-diameter
balls. Inside the mill, which has an energy consumption of 1 MW and a capacity of 50 tons/hr, the
waste is ground and mixed with the sludge. This mixture is then discharged via a trommel screen
III-15
-------
oj
&
•kj
&
Q
Q
OJ
£
s^
O
co
o
**
5
Q)
•s
•5
1
o
p.
.co
"S
0
o
CO
••v
o
.
"— ' Q.
QC *"* ro
Si0
a.
O)
CD .£
to to to
-3 10 CD
•due
CD O — ,
oc o -1
O.
>.
O)
° c
1-t
•go
CD
^
a.
3
e
CD
CO
c
CO
0.
r^
~T
>
73
C
CO
J:
B
_i
CD
.C
D.
4->
to
a
+j
to
in
CO
K
#
©
£
a.
*-•
to
2.
Q
a.
1-
to
?>
^
a
to
CN
O
•a
c
CD
>
^
CD
CD
>-
ID
CD
CO
CD
e
CO
Z
O
~z.
Dried sludge either
burnt in suspension
o
Hot gases are
exclusively used
#
in
o
x0.38@100
>-
8
CN
l-~
•*
X
CN
--. w
JC o
0 JI
s °
a> Z
a
•
I- CO
K a
~ C/5
£ Q.
Q)
W) "D
"D CD
a> >
!l
O £
cb
c
'>.
73
CD
O)
•o
_3
CO
dryer problems.
O
^
Test runs commenced
during 1979.
Equipment modifica-
tions in 1980 and
1982. Full design
capacity not yet
accomplished.
ra
'to
Sg§
O CN
O -s c
O (& co
*"!.!
X CO ^
CN S£
"*
CN
o
in
CN
©
00
o
t-^
X
CN
O
§
m
^s:
oo£
x£
CN
73
C
CO
CD
Q.
0
O
o
00
en
*~
~2L
5
4,
3
3
Q
O
2
Demonstration plant,
now abandoned and
dismantled.
CD
c
o
z
00
CN
CD
x0.42@100
«-
0
in
CN
X
«~
c
^ 0
o9 to
-^ m
CJ CO
CO [J
m
,—
r~
en
»—
I
O
c
~
C
(0
u_
O
2
Circular batch-fed in-
cinerators, only one
adapted to sludge.
1967 wet scrubber in-
stalled, closed 1975.
0 O
Only hot gases a
exclusively used
sludge drying.
CD
o
CD
o
o
i;
co
oo
CN
6
^_
o
o
CO
CO
CD
X
CN
73
C
O
E
>-
CO
o:
UJ
U
00
CD
i^
in
en
H
O
•o
0
"^
CD
X
o
•^
A third incinerator
unit was to be built
to eliminate need for
CD
a>4?
m 3
CO «
CD •-
to O
CO t-
01 tj
m CD
3 to .
^ 3 OJ
|tl
X O 73
CN
CD
in
oo
@
CN
•^•
CD
X
^~
in
CN
CN
OT
co
•*
X
CN
£.
D)
3
n
to
^
i£
m
CD
en
T—
<
^
V
_*
o
>-
o
supplemental oil firing.
Closed 1976 because
landfilling was cheaper
than upgrading.
Ill-16